Compare commits

..
Author SHA1 Message Date
JackDoan 04e80f0dcf improvements to config parsing 2026-10-01 10:18:30 -05:00
JackDoan ed8ae57fee checkpt 2026-09-30 15:06:35 -05:00
JackDoan e99e48cca7 checkpt 2026-09-30 14:57:19 -05:00
JackDoan 027c728702 checkpt 2026-09-30 14:52:14 -05:00
JackDoan 1cfed59214 checkpt 2026-09-30 14:46:44 -05:00
JackDoan c4b9df20ac u64->u8 2026-09-30 14:13:21 -05:00
JackDoan 92bad7460b checkpt 2026-09-30 14:04:05 -05:00
JackDoan cac7586c0b checkpt 2026-09-30 14:03:26 -05:00
JackDoan 9aab46c1ad no bart for cidr 2026-09-30 12:24:01 -05:00
JackDoan f3b76703ae maybe better 2026-09-29 19:51:08 -05:00
JackDoan fc0dbfec4c maybe better 2026-09-29 19:49:27 -05:00
JackDoan ca69a0613b memory oh baby 2026-09-29 19:38:53 -05:00
JackDoan 84a26846f4 refactor checkpt 2026-09-29 18:41:01 -05:00
JackDoan 317388e37d refactor checkpt 2026-09-29 18:36:48 -05:00
JackDoan 06fd191367 stuff 2026-09-29 18:04:53 -05:00
JackDoan 3f392883ca stuff 2026-09-29 17:52:13 -05:00
JackDoan 30cf2bbe6c Don't check port any firewall rules twice for packets without ports 2026-09-29 17:38:49 -05:00
JackDoan 11b04517f3 Fold proto any firewall rules into each protocol's rules so a packet checks one table 2026-09-29 17:38:19 -05:00
JackDoan 610ba57105 Allow firewall rules on any IP protocol number 2026-09-29 17:24:09 -05:00
JackDoan 9e96894bd7 Pass firewall packets by pointer through conntrack and rule matching 2026-09-29 17:18:11 -05:00
JackDoan 14f80450fb allow more protos to have ports parsed 2026-09-29 15:51:59 -05:00
JackDoan 641d1fb703 Clarify which traffic port rules match 2026-09-29 13:17:21 -05:00
JackDoan 11a4298c36 Only read IPv4 ports from TCP and UDP 2026-09-29 13:02:30 -05:00
163 changed files with 4027 additions and 16583 deletions

No files matched your search

-2
View File
@@ -1,2 +0,0 @@
# Goldens are compared byte for byte, so they must come out of git the way they went in.
*.golden -text
+3 -3
View File
@@ -14,7 +14,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: Build - name: Build
@@ -40,7 +40,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: Build - name: Build
@@ -83,7 +83,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
# GitHub holds ARNs, not credentials, and ARNs outlive a rotation # GitHub holds ARNs, not credentials, and ARNs outlive a rotation
+3 -3
View File
@@ -34,7 +34,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: add hashicorp source - name: add hashicorp source
@@ -66,7 +66,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: add hashicorp source - name: add hashicorp source
@@ -92,7 +92,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
# WSL2 + Ubuntu so the smoke can run a real linux peer with its own # WSL2 + Ubuntu so the smoke can run a real linux peer with its own
+2 -2
View File
@@ -22,7 +22,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: Smoke Docker - name: Smoke Docker
@@ -51,7 +51,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: build - name: build
+4 -4
View File
@@ -22,7 +22,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: Install goimports - name: Install goimports
@@ -42,7 +42,7 @@ jobs:
- name: golangci-lint - name: golangci-lint
uses: golangci/golangci-lint-action@v9 uses: golangci/golangci-lint-action@v9
with: with:
version: v2.14 version: v2.12
test: test:
name: Test ${{ matrix.name }} name: Test ${{ matrix.name }}
@@ -87,7 +87,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: Build - name: Build
@@ -132,7 +132,7 @@ jobs:
- uses: actions/setup-go@v7 - uses: actions/setup-go@v7
with: with:
go-version: '1.27' go-version: '1.26'
check-latest: true check-latest: true
- name: Build ${{ matrix.name }} - name: Build ${{ matrix.name }}
+12 -14
View File
@@ -7,24 +7,22 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
## [Unreleased] ## [Unreleased]
### Added ### Changed
- A control socket that provides access to all of the ssh debug server's commands along with support - Firewall rules are held as a flat list with a bitset of rule ids per port, instead of a tree of maps copied
for running `nebula ctl <command>` as a client from the CLI. into every port of a range. A rule like `port: 1024-65535` now costs kilobytes rather than tens of megabytes,
and matching is faster on every path.
### Removed - A firewall rule's `proto` can be `udplite`, `dccp`, `sctp`, `icmpv6`, or any IP protocol number. A rule with
a port or port range on a protocol that has no ports, such as `proto: 47`, is rejected rather than silently
- The SSH debug server's `start-cpu-profile`, `stop-cpu-profile`, `save-heap-profile`, `save-mutex-profile`, and never matching. `port: any` and `port: fragment` apply to every protocol.
`mutex-profile-fraction` commands, and `sshd.sandbox_dir`, which is now ignored if set. - An ICMP non-first fragment is matched against `port: fragment` rules like a fragment of any other protocol.
It was only matched against `port: any` rules before.
### Fixed ### Fixed
- Signing a v2 certificate with an empty name, a name longer than 253 bytes, or an empty group now fails. - A firewall rule with `group: []`, `groups:` with no value, or a `groups` entry that isn't a string, such as
These certificates were signed successfully but could not be loaded. `groups: [1, 2]`, no longer crashes Nebula on startup or reload. An empty `group` array is rejected, a null
- `create-tunnel -address <ip:port>` panicked on a nil RemoteList and, over the ssh debug server, took nebula `groups` means no groups, and non-string entries are read as strings.
down with it. A hostinfo returned by StartHandshake had no RemoteList until the handshake loop's first
pass attached one, and `create-tunnel -address` set the remote before that. The list is now attached when
the handshake starts, so a pending hostinfo never exists without one.
## [1.11.1] - 2026-08-21 ## [1.11.1] - 2026-08-21
+1 -10
View File
@@ -389,16 +389,7 @@ func (c *certificateV2) fromTBSCertificate(t *TBSCertificate) error {
} }
func (c *certificateV2) validate() error { func (c *certificateV2) validate() error {
// Mirror the name and group rules in unmarshalDetails so signing fails instead of producing an unparseable certificate // Empty names are allowed
if c.details.name == "" || len(c.details.name) > MaxNameLength {
return NewErrInvalidCertificateProperties("name must be between 1 and %d bytes", MaxNameLength)
}
for _, group := range c.details.groups {
if group == "" {
return NewErrInvalidCertificateProperties("groups may not be empty")
}
}
if len(c.publicKey) == 0 { if len(c.publicKey) == 0 {
return ErrInvalidPublicKey return ErrInvalidPublicKey
-45
View File
@@ -6,7 +6,6 @@ import (
"crypto/elliptic" "crypto/elliptic"
"crypto/rand" "crypto/rand"
"net/netip" "net/netip"
"strings"
"testing" "testing"
"time" "time"
@@ -136,47 +135,3 @@ func TestCertificate_SignP256_AlwaysNormalized(t *testing.T) {
assert.True(t, normie) assert.True(t, normie)
} }
} }
func TestCertificateV2_SignRejectsUnparseable(t *testing.T) {
_, priv, err := ed25519.GenerateKey(rand.Reader)
require.NoError(t, err)
var invalid *ErrInvalidCertificateProperties
newTBS := func() TBSCertificate {
return TBSCertificate{
Version: Version2,
Name: "testing",
Networks: []netip.Prefix{mustParsePrefixUnmapped("10.1.1.1/24")},
Groups: []string{"test-group1"},
NotBefore: time.Now().Add(-time.Minute),
NotAfter: time.Now().Add(time.Minute),
PublicKey: []byte("1234567890abcedfghij1234567890ab"),
Curve: Curve_CURVE25519,
IsCA: true,
}
}
tbs := newTBS()
tbs.Name = strings.Repeat("a", MaxNameLength)
c, err := tbs.Sign(nil, Curve_CURVE25519, priv)
require.NoError(t, err)
b, err := c.Marshal()
require.NoError(t, err)
_, err = unmarshalCertificateV2(b, nil, Curve_CURVE25519)
require.NoError(t, err)
tbs = newTBS()
tbs.Name = strings.Repeat("a", MaxNameLength+1)
_, err = tbs.Sign(nil, Curve_CURVE25519, priv)
require.ErrorAs(t, err, &invalid)
tbs = newTBS()
tbs.Name = ""
_, err = tbs.Sign(nil, Curve_CURVE25519, priv)
require.ErrorAs(t, err, &invalid)
tbs = newTBS()
tbs.Groups = []string{"test-group1", ""}
_, err = tbs.Sign(nil, Curve_CURVE25519, priv)
require.ErrorAs(t, err, &invalid)
}
-420
View File
@@ -1,420 +0,0 @@
package ctl
import (
"bytes"
"encoding/json"
"errors"
"flag"
"fmt"
"io"
"log/slog"
"os"
"strings"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/localapi"
)
// Word is the argument that selects the control client instead of the daemon.
const Word = "ctl"
// Subcommand handles the one bare word a daemon binary accepts before its flags. `nebula
// ctl <command>` runs the client. Any other first argument that is not a flag is a
// mistyped subcommand, reported with a usage line: the daemon's flag parser would ignore
// the stray word and start nebula, so `nebula clt` would bring up a second daemon
// instead of reporting a typo. It returns the exit code and true when the process is
// done, and false when argv is the daemon's to parse. Both daemon binaries, cmd/nebula
// and cmd/nebula-service, call it before anything else, so each has a nebula ctl and
// neither has the hazard. A word that follows the daemon's flags instead is Trailing's.
func Subcommand(argv []string) (code int, done bool) {
if len(argv) < 2 || strings.HasPrefix(argv[1], "-") {
return 0, false
}
if argv[1] == Word {
return Run(argv[2:]), true
}
return unknownCommand(argv[1]), true
}
// Trailing handles the words the daemon's flag parser leaves unparsed: flag.Parse stops
// at the first argument that is not a flag and hands the rest back. rest is flag.Args()
// after the daemon has parsed, and configPath is the config the daemon's -config named,
// "" when it did not. Both daemon binaries call it right after flag.Parse, so that
// `nebula -config x ctl identity` runs the client and `nebula -config x clt` reports a
// typo, as Subcommand does for a word at the front of the line; a daemon that ignored
// the words would start in their place.
//
// A leading ctl word runs the client on what follows, with the daemon's -config carried
// over: an operator who typed it meant the daemon running on that config, and the client
// derives the same socket from it. Any other word is a mistake, reported as Subcommand
// reports one.
func Trailing(rest []string, configPath string) (code int, done bool) {
if len(rest) == 0 {
return 0, false
}
if rest[0] != Word {
return unknownCommand(rest[0]), true
}
args := rest[1:]
if configPath != "" {
// First, so a -config on the ctl side of the line still wins: Parse applies
// flags in order and the last value set is the one read.
args = append([]string{"-" + commands.FlagConfig, configPath}, args...)
}
return Run(args), true
}
// unknownCommand reports a bare word that is neither a flag nor ctl, with the one usage
// line that says what the two forms of the command line are.
func unknownCommand(word string) int {
fmt.Fprintf(os.Stderr, "nebula: unknown command %q\n\n", word)
fmt.Fprintf(os.Stderr, "Run nebula with flags to start it, or `nebula %s <command>` to control one that is already running.\n", Word)
return 1
}
// Run implements `nebula ctl <command> [args]`
//
// args is everything after the `ctl` word.
func Run(args []string) int {
fs := flag.NewFlagSet("nebula ctl", flag.ContinueOnError)
fs.SetOutput(os.Stderr)
configPath := fs.String(commands.FlagConfig, "", "Path to the config of the running nebula, used to locate its control socket")
socketPath := fs.String(commands.FlagSocket, "", "Path to the control socket, instead of the one the config implies")
fs.Usage = func() {
fmt.Fprintf(os.Stderr, "Usage: nebula ctl [flags] <command> [command flags]\n\n")
fmt.Fprintf(os.Stderr, "Inspects and controls a nebula already running on this machine.\n")
fmt.Fprintf(os.Stderr, "Run it as the user who ran nebula, or as a member of the group %s names.\n\n", localapi.CtlGroupKey)
fmt.Fprintf(os.Stderr, "Flags, which may come before or after the command:\n")
fs.PrintDefaults()
fmt.Fprintf(os.Stderr, "\nRun `nebula ctl help` for the available commands.\n")
}
own, cmdArgs := splitLine(fs, args)
if err := fs.Parse(own); err != nil {
// fs has already printed the usage, or the problem. Asking for help is not a
// failure, and a script that chains on `nebula ctl -h` should not see one.
if errors.Is(err, flag.ErrHelp) {
return 0
}
return 1
}
if len(cmdArgs) == 0 {
cmdArgs = []string{"help"}
}
path, err := resolveSocket(*socketPath, *configPath)
if err != nil {
fmt.Fprintf(os.Stderr, "%s\n", err)
return 1
}
client := &localapi.Client{Path: path}
// Ask the daemon what it serves before parsing anything. The flags belong to the
// command, which lives over there, so this is what lets the CLI keep its usual shape
// while the wire protocol stays a named-parameter API rather than a shipped argv.
schema, err := client.Describe()
if err != nil {
// A daemon that answered was reached, and what it said, or that it could not be
// read, is the whole message.
var refused *localapi.DaemonError
var unread *localapi.ReplyError
if errors.As(err, &refused) || errors.As(err, &unread) {
fmt.Fprintf(os.Stderr, "nebula: %s\n", err)
return 1
}
// Otherwise the operating system said why, and it is passed on as it came: a
// permission denied or a missing socket says more about what to do than a
// guess at the operator's setup would.
fmt.Fprintf(os.Stderr, "nebula: could not reach the control socket at %s: %s\n", path, err)
return 1
}
req, printedHelp, err := parseCommand(schema, cmdArgs)
if err != nil {
fmt.Fprintf(os.Stderr, "%s\n", err)
return 1
}
if printedHelp {
return 0
}
res, err := client.Invoke(string(req.name), req.Params, req.Args, wireFormat(req.Format))
if err != nil {
fmt.Fprintf(os.Stderr, "nebula: %s\n", err)
return 1
}
out := output(res, req.Format)
if out != "" {
fmt.Fprint(os.Stdout, out)
if !strings.HasSuffix(out, "\n") {
fmt.Fprintln(os.Stdout)
}
}
// Error set is the one failure signal, whether the request never ran or the command
// ran and refused. A refusal asked for in JSON has already gone to stdout as the
// {ok, reason} object, and is not said again on stderr; the exit status carries it.
if res.Error == "" {
return 0
}
if len(res.Result) == 0 && len(res.Text) == 0 {
msg := res.Error
if !strings.HasSuffix(msg, "\n") {
msg += "\n"
}
fmt.Fprint(os.Stderr, msg)
}
return 1
}
// output renders what the daemon answered for the format the line asked for. A text
// request is answered in Text with the bytes the command wrote, printed as they are; a
// JSON request is answered in Result with the value itself, printed as it came, since a
// value that happens to be a string is still JSON to the script reading it. Indentation
// is the client's business: whitespace does not survive a JSON envelope, and the daemon
// has no reason to care how its answer is laid out. The bytes are re-indented rather
// than decoded and encoded again, which would sort the keys and round the numbers.
func output(res *localapi.Response, format commands.Format) string {
switch format {
case commands.FormatJSON:
return string(res.Result)
case commands.FormatPrettyJSON:
var buf bytes.Buffer
if err := json.Indent(&buf, res.Result, "", " "); err != nil {
return string(res.Result)
}
return buf.String()
default:
return string(res.Text)
}
}
// ctlRequest is one parsed nebula ctl command line: the command's name, which only the
// daemon can resolve, and the line as every surface parses it. The Request's Command is
// nil here, since the CLI holds a schema rather than the command, and its Format is
// FormatUnset when the line named none, which to nebula ctl means text: its default
// differs from the API's.
type ctlRequest struct {
name commands.Name
commands.Request
}
// wireFormat is what a request asks the daemon for. Pretty is the client's business,
// since whitespace does not survive a JSON envelope, so it travels as json. No
// preference is text, nebula ctl's default: a person gets the human rendering wherever
// one exists.
func wireFormat(f commands.Format) string {
switch f {
case commands.FormatJSON, commands.FormatPrettyJSON:
return "json"
default:
return "text"
}
}
// parseCommand turns the user's command line into a ctlRequest, using the schema the
// daemon supplied. The line is parsed with the same rules the daemon applies to an ssh
// line, through ParseArgs, and ParseLine reads what was set into the Request to send, so
// a line means the same on both.
//
// printedHelp reports that help was written for the command instead, and there is
// nothing to send.
func parseCommand(schema []commands.CommandSchema, argv []string) (req ctlRequest, printedHelp bool, err error) {
name := commands.Name(argv[0])
var found *commands.CommandSchema
for i := range schema {
if schema[i].Name == name {
found = &schema[i]
break
}
}
if found == nil {
// Let the daemon answer. It knows whether this is a typo or a command it holds
// back, and its message for the second case is better than anything guessable
// from the schema alone.
return ctlRequest{name: name, Request: commands.Request{Args: argv[1:]}}, false, nil
}
// The flag package's own report is a bare message followed by a usage dump in a layout
// that does not match WriteCommandHelp, so it is silenced and the error is reported
// once, below, with a pointer at the help this command actually prints.
fs := flag.NewFlagSet(string(name), flag.ContinueOnError)
fs.SetOutput(io.Discard)
for _, f := range found.Flags {
if f.Bool {
fs.Bool(f.Name, f.Default == "true", f.Usage)
continue
}
fs.String(f.Name, f.Default, f.Usage)
}
operands, err := commands.ParseArgs(fs, argv[1:])
if err != nil {
if errors.Is(err, flag.ErrHelp) {
// The same rendering the daemon uses for `help <command>`, so -h and help agree.
_ = commands.WriteCommandHelp(*found, os.Stdout)
return ctlRequest{}, true, nil
}
// Say what flag did: an unknown flag, a missing value and a malformed value are
// different mistakes, and "does not take" would be wrong for two of them.
return ctlRequest{}, false, fmt.Errorf("nebula: %s: %s (see 'nebula ctl %s -h')", name, err, name)
}
// Only what the user actually set travels. A parameter left alone is the daemon's
// default to apply, not the client's to assert, and the presentation flags are the
// client's to interpret, never sent.
//
// Where the command has no human form the daemon says so, and only once the command
// has run: a refusal ("No vpn address was provided") always has a human form, so
// telling the user to pass -json before finding out whether the command would even
// accept the line sends them to fix the wrong thing. That is why the schema does not
// say which commands have a human form: nothing should decide on it before running.
return ctlRequest{name: name, Request: commands.ParseLine(fs, operands)}, false, nil
}
// splitLine divides a nebula ctl line into the client's own flags, for fs to parse, and
// the command line to send, which begins with the command word. The client's flags may
// come before the command or after it: `nebula ctl -config x identity` and
// `nebula ctl identity -config x` mean the same, so an operator who has typed a command
// and then remembers which daemon they meant does not have to retype the line.
//
// Before the command word every flag is the client's, whatever it is, so an unknown one
// is reported by fs, and -h is the client's usage. After it only the
// flags fs defines are lifted out, and only those: the command's own flags are unknown
// here, since they come from the daemon the line has yet to name, and -h after the
// command asks for that command's help. Which command flags take a value is unknown too,
// so the client's names are reserved in the commands package and no command can give
// one a value that would be lifted from here. A "--" before the command is consumed, as
// fs.Parse would have; one after it belongs to the command line and ends the lifting,
// so an operand spelled like a client flag can still be passed.
func splitLine(fs *flag.FlagSet, args []string) (own, cmd []string) {
seenCommand := false
for i := 0; i < len(args); i++ {
a := args[i]
if a == "--" {
if seenCommand {
return own, append(cmd, args[i:]...)
}
seenCommand = true
continue
}
if !seenCommand {
if len(a) < 2 || a[0] != '-' {
seenCommand = true
cmd = append(cmd, a)
continue
}
own = append(own, a)
if f, hasValue := commands.LookupArg(fs, a); f != nil && !hasValue && !commands.IsBoolFlag(f) && i+1 < len(args) {
i++
own = append(own, args[i])
}
continue
}
if len(a) >= 2 && a[0] == '-' {
// The same reading of the argument ParseArgs makes, so a word this lifts as a
// client flag's value is one the daemon's parser would have read the same way.
if f, hasValue := commands.LookupArg(fs, a); f != nil {
own = append(own, a)
if !hasValue && !commands.IsBoolFlag(f) && i+1 < len(args) {
i++
own = append(own, args[i])
}
continue
}
}
cmd = append(cmd, a)
}
return own, cmd
}
// resolveSocket decides which socket to talk to. It never searches: the answer is the
// socket the operator named, or the one the config implies, and nothing else.
//
// An explicit flag wins. Otherwise the config is read, the one the operator named or the
// one found next to the executable, and the path is the socket that config implies: its
// local_api.control.socket if set, else the same derivation the daemon applies.
func resolveSocket(socketFlag, configFlag string) (string, error) {
if socketFlag != "" {
return socketFlag, nil
}
return pathFromConfig(configFlag)
}
// pathFromConfig returns the socket path the daemon would be serving: the one
// local_api.control.socket names in the config, so an operator who moved the socket does
// not have to repeat themselves on every invocation, else the default.
//
// A config the operator did not name is a guess: config.DefaultPath looks next to the
// executable, and a package that installs a default config there while systemd starts
// the daemon with another will make it the wrong guess. It is still the only answer
// that involves no searching. A guess that cannot be read is not an error, it is no
// guess: a member of the socket's group cannot read a root daemon's config, and is
// answered by the default path, where the daemon serves unless that config moved it.
// A config the operator named is theirs to make readable.
func pathFromConfig(configFlag string) (string, error) {
return socketFromConfigFile(configFlag, configFlag != "")
}
func socketFromConfigFile(configPath string, named bool) (string, error) {
if !named {
p, err := config.DefaultPath()
if err != nil {
return localapi.DefaultSocketPath, nil
}
configPath = p
}
// The config is loaded only to find the socket path. Loading logs lines that belong
// to the daemon, not to someone running a read only command, and mixing them into
// the CLI's output would corrupt a -json result, so they are discarded.
c := config.NewC(slog.New(slog.DiscardHandler))
if err := c.Load(configPath); err != nil {
if !named {
return localapi.DefaultSocketPath, nil
}
// The operating system's error, as it came: permission denied on a root daemon's
// config says what to do.
return "", fmt.Errorf("nebula: could not read %s: %w", configPath, err)
}
// A daemon running this config serves no socket; dialing the path would only say
// that nothing is there.
if !localapi.ControlEnabled(c) {
return "", fmt.Errorf("nebula: the control socket is disabled in %s by %s", configPath, localapi.CtlEnabledKey)
}
// The same call the daemon makes, so the two agree on the path and on what makes a
// path unusable.
p, err := localapi.SocketPathFromConfig(c)
if err != nil {
return "", fmt.Errorf("nebula: %w", err)
}
return p, nil
}
-294
View File
@@ -1,294 +0,0 @@
package ctl
import (
"encoding/json"
"flag"
"os"
"path/filepath"
"runtime"
"strings"
"testing"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/localapi"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestResolveSocketReportsAnUnreadableConfig: an operator who named a config must hear
// that it could not be read, rather than have the command steered anywhere else.
func TestResolveSocketReportsAnUnreadableConfig(t *testing.T) {
missing := filepath.Join(t.TempDir(), "typo.yml")
_, err := resolveSocket("", missing)
require.Error(t, err)
assert.Contains(t, err.Error(), missing)
}
// TestResolveSocketGuessedConfigNeedNotBeReadable: a config nobody named is a guess, and
// one that cannot be read, a root daemon's config read by a member of the socket's
// group, means the default path rather than an error.
func TestResolveSocketGuessedConfigNeedNotBeReadable(t *testing.T) {
missing := filepath.Join(t.TempDir(), "config.yml")
p, err := socketFromConfigFile(missing, false)
require.NoError(t, err)
assert.Equal(t, localapi.DefaultSocketPath, p)
_, err = socketFromConfigFile(missing, true)
require.ErrorContains(t, err, missing)
}
// nowhereSocket is a socket name nothing listens on, spelled the way the platform
// requires: a full pipe name on windows, an absolute path elsewhere.
var nowhereSocket = map[bool]string{true: `\\.\pipe\nebula-nowhere`, false: "/nowhere/ctl.sock"}[runtime.GOOS == "windows"]
// namedSocketConfig is a config that names nowhereSocket as the control socket. Single
// quotes keep a pipe name's backslashes literal in YAML.
var namedSocketConfig = "local_api:\n control:\n socket: '" + nowhereSocket + "'\n"
// TestResolveSocketTrustsANamedConfig: a config the operator named yields the socket it
// implies even when nothing is listening there, so a dial failure names that socket.
func TestResolveSocketTrustsANamedConfig(t *testing.T) {
cfg := filepath.Join(t.TempDir(), "config.yml")
require.NoError(t, os.WriteFile(cfg, []byte(namedSocketConfig), 0600))
p, err := resolveSocket("", cfg)
require.NoError(t, err)
assert.Equal(t, nowhereSocket, p)
}
// TestResolveSocketSaysTheSocketIsDisabled: a config that turns the control socket off
// is answered with that, not with a dial to a path nothing serves. -socket still wins.
func TestResolveSocketSaysTheSocketIsDisabled(t *testing.T) {
cfg := filepath.Join(t.TempDir(), "config.yml")
require.NoError(t, os.WriteFile(cfg, []byte("local_api:\n control:\n enabled: false\n"), 0600))
_, err := resolveSocket("", cfg)
require.ErrorContains(t, err, "the control socket is disabled in "+cfg+" by local_api.control.enabled")
p, err := resolveSocket("/explicit.sock", cfg)
require.NoError(t, err)
assert.Equal(t, "/explicit.sock", p)
}
// TestResolveSocketUsesTheDefaultWithoutAnOverride: a config that does not name the
// socket means the default path, the same one the daemon serves with that config, and
// no config at all means the default too, since that is where a daemon started without
// an override serves. Nothing is searched for.
func TestResolveSocketUsesTheDefaultWithoutAnOverride(t *testing.T) {
cfg := filepath.Join(t.TempDir(), "config.yml")
require.NoError(t, os.WriteFile(cfg, []byte("static_host_map: {}\n"), 0600))
p, err := resolveSocket("", cfg)
require.NoError(t, err)
assert.Equal(t, localapi.DefaultSocketPath, p)
// The test binary has no config next to it.
p, err = resolveSocket("", "")
require.NoError(t, err)
assert.Equal(t, localapi.DefaultSocketPath, p)
}
func TestResolveSocketFlagWins(t *testing.T) {
p, err := resolveSocket("/explicit.sock", "/does/not/matter.yml")
require.NoError(t, err)
assert.Equal(t, "/explicit.sock", p)
}
// TestParseCommandReportsWhatFlagSaid covers the three ways a flag line goes wrong, which
// are different mistakes: an unknown flag, a valid flag with a missing value, and a
// valid flag with a value it cannot parse.
func TestParseCommandReportsWhatFlagSaid(t *testing.T) {
schema := []commands.CommandSchema{{
Name: "change-remote",
Flags: []commands.FlagSchema{
{Name: "address"},
{Name: "by-index", Bool: true},
},
}}
cases := map[string]struct {
argv []string
want string
}{
"missing value": {[]string{"change-remote", "-address"}, "flag needs an argument: -address"},
"bad bool": {[]string{"change-remote", "-by-index=maybe"}, "invalid boolean value \"maybe\" for -by-index"},
"unknown flag": {[]string{"change-remote", "-nope"}, "flag provided but not defined: -nope"},
}
for name, tc := range cases {
t.Run(name, func(t *testing.T) {
_, _, err := parseCommand(schema, tc.argv)
require.Error(t, err)
require.ErrorContains(t, err, "nebula: change-remote: "+tc.want)
require.ErrorContains(t, err, "see 'nebula ctl change-remote -h'")
assert.NotContains(t, err.Error(), "does not take")
})
}
}
// TestParseCommandOffPresentationFlagIsNoPreference: the daemon treats -json=false as
// no preference, so nebula ctl must too, or the same line renders differently on the
// two surfaces. Only a true value asks for JSON, and -pretty implies it.
func TestParseCommandOffPresentationFlagIsNoPreference(t *testing.T) {
schema := []commands.CommandSchema{{
Name: "device-info",
Flags: []commands.FlagSchema{
{Name: commands.FlagJSON, Bool: true, Default: "false"},
{Name: commands.FlagPretty, Bool: true, Default: "false"},
},
}}
for _, tc := range []struct {
argv []string
want commands.Format
wire string
}{
{[]string{"device-info"}, commands.FormatUnset, "text"},
{[]string{"device-info", "-json=false"}, commands.FormatUnset, "text"},
{[]string{"device-info", "-json"}, commands.FormatJSON, "json"},
{[]string{"device-info", "-pretty"}, commands.FormatPrettyJSON, "json"},
} {
t.Run(strings.Join(tc.argv[1:], " "), func(t *testing.T) {
req, printedHelp, err := parseCommand(schema, tc.argv)
require.NoError(t, err)
assert.False(t, printedHelp)
assert.Equal(t, tc.want, req.Format)
assert.Equal(t, tc.wire, wireFormat(req.Format))
assert.Nil(t, req.Params, "presentation flags are never sent as parameters")
})
}
}
// TestParseCommandReadsTheLineAsTheDaemonDoes: nebula ctl parses the line with the same
// rules as the daemon, so an operand after "--" reaches the command on both and a
// negative number before it is the flag it would be at an ssh prompt.
func TestParseCommandReadsTheLineAsTheDaemonDoes(t *testing.T) {
schema := []commands.CommandSchema{{Name: "set"}}
req, printedHelp, err := parseCommand(schema, []string{"set", "--", "-1"})
require.NoError(t, err)
assert.False(t, printedHelp)
assert.Equal(t, commands.Name("set"), req.name)
assert.Nil(t, req.Params)
assert.Equal(t, []string{"-1"}, req.Args)
_, _, err = parseCommand(schema, []string{"set", "-1"})
require.ErrorContains(t, err, "flag provided but not defined: -1")
}
// TestOutputRendersForTheFormatAsked: what the daemon answered is read by the format the
// line asked for, never by sniffing the bytes. A text answer is the command's rendering,
// printed as it came; a JSON answer whose value is a string stays quoted, since to the
// script reading it that string is the JSON.
func TestOutputRendersForTheFormatAsked(t *testing.T) {
text := &localapi.Response{Text: []byte("Log level is: info\n")}
assert.Equal(t, "Log level is: info\n", output(text, commands.FormatUnset))
assert.Equal(t, "Log level is: info\n", output(text, commands.FormatText))
str := &localapi.Response{Result: json.RawMessage(`"some string"`)}
assert.Equal(t, `"some string"`, output(str, commands.FormatJSON))
assert.Equal(t, `"some string"`, output(str, commands.FormatPrettyJSON))
// Keys stay in the order the daemon wrote them and a uint64 past 2^53 is not rounded,
// which decoding into `any` and encoding again would do.
obj := &localapi.Response{Result: json.RawMessage(`{"z":18446744073709551615,"a":{"y":[1,2],"b":true}}`)}
//nolint:testifylint // the layout is what is under test, which JSONEq would not see
assert.Equal(t, `{"z":18446744073709551615,"a":{"y":[1,2],"b":true}}`, output(obj, commands.FormatJSON))
//nolint:testifylint // as above
assert.Equal(t, "{\n \"z\": 18446744073709551615,\n \"a\": {\n \"y\": [\n 1,\n 2\n ],\n \"b\": true\n }\n}", output(obj, commands.FormatPrettyJSON))
bad := &localapi.Response{Result: json.RawMessage(`{"a":`)}
assert.Equal(t, `{"a":`, output(bad, commands.FormatPrettyJSON), "what cannot be indented is printed as it came")
none := &localapi.Response{}
assert.Empty(t, output(none, commands.FormatJSON))
assert.Empty(t, output(none, commands.FormatUnset))
}
// TestSplitLineLiftsOwnFlagsFromAnywhere: -config and -socket are the client's wherever
// they appear, before the command or after it, in either spelling, so the two orders
// mean the same line. Everything else after the command word stays in order for the
// command's own parse, including flags and operands that happen to look like values.
func TestSplitLineLiftsOwnFlagsFromAnywhere(t *testing.T) {
newFS := func() *flag.FlagSet {
fs := flag.NewFlagSet("nebula ctl", flag.ContinueOnError)
fs.String(commands.FlagConfig, "", "")
fs.String(commands.FlagSocket, "", "")
return fs
}
for _, tc := range []struct {
args []string
wantOwn []string
wantCmd []string
}{
{nil, nil, nil},
{[]string{"identity"}, nil, []string{"identity"}},
{[]string{"-config", "x", "identity"}, []string{"-config", "x"}, []string{"identity"}},
{[]string{"identity", "-config", "x"}, []string{"-config", "x"}, []string{"identity"}},
{[]string{"identity", "-config=x"}, []string{"-config=x"}, []string{"identity"}},
{[]string{"identity", "--socket", "s"}, []string{"--socket", "s"}, []string{"identity"}},
{[]string{"-socket", "s", "identity", "-config", "x"}, []string{"-socket", "s", "-config", "x"}, []string{"identity"}},
// The command's flags and operands are left where they are, around the lifted ones.
{[]string{"print-tunnel", "10.0.0.1", "-config", "x", "-json"}, []string{"-config", "x"}, []string{"print-tunnel", "10.0.0.1", "-json"}},
{[]string{"change-remote", "-address", "1.2.3.4:5", "-socket", "s", "10.0.0.1"}, []string{"-socket", "s"}, []string{"change-remote", "-address", "1.2.3.4:5", "10.0.0.1"}},
// -h before the command is the client's usage; after it, the command's help.
{[]string{"-h"}, []string{"-h"}, nil},
{[]string{"identity", "-h"}, nil, []string{"identity", "-h"}},
// An unknown flag before the command is the client's to report, and does not
// swallow the command word as its value.
{[]string{"-nope", "identity"}, []string{"-nope"}, []string{"identity"}},
// A -config with nothing after it is left for fs.Parse to report.
{[]string{"identity", "-config"}, []string{"-config"}, []string{"identity"}},
// "--" before the command is consumed, as fs.Parse consumed it; after the command
// it belongs to the line and ends the lifting.
{[]string{"--", "identity", "-config", "x"}, []string{"-config", "x"}, []string{"identity"}},
{[]string{"identity", "--", "-config", "x"}, nil, []string{"identity", "--", "-config", "x"}},
} {
t.Run(strings.Join(tc.args, " "), func(t *testing.T) {
own, cmd := splitLine(newFS(), tc.args)
assert.Equal(t, tc.wantOwn, own)
assert.Equal(t, tc.wantCmd, cmd)
})
}
}
// TestTrailingHandlesWordsAfterTheDaemonsFlags: what flag.Parse leaves behind is never
// a daemon to start. A leading ctl runs the client on the rest, with the daemon's
// -config carried over; anything else is a mistyped subcommand. Only -h is run here,
// since the client's help needs no daemon: it proves the words reach Run and that a
// carried -config is accepted by it.
func TestTrailingHandlesWordsAfterTheDaemonsFlags(t *testing.T) {
code, done := Trailing(nil, "")
assert.False(t, done)
assert.Equal(t, 0, code)
code, done = Trailing([]string{"clt"}, "")
assert.True(t, done)
assert.Equal(t, 1, code)
code, done = Trailing([]string{Word, "-h"}, "")
assert.True(t, done)
assert.Equal(t, 0, code)
code, done = Trailing([]string{Word, "-h"}, "/etc/nebula/config.yml")
assert.True(t, done)
assert.Equal(t, 0, code)
}
// TestSubcommandOnlyReadsTheFirstWord: a flag first means the line is the daemon's to
// parse, and whatever follows the flags is Trailing's to judge.
func TestSubcommandOnlyReadsTheFirstWord(t *testing.T) {
_, done := Subcommand([]string{"nebula", "-config", "x", Word})
assert.False(t, done)
code, done := Subcommand([]string{"nebula", "clt"})
assert.True(t, done)
assert.Equal(t, 1, code)
code, done = Subcommand([]string{"nebula", Word, "-h"})
assert.True(t, done)
assert.Equal(t, 0, code)
}
-42
View File
@@ -1,42 +0,0 @@
//go:build !windows
package ctl
import (
"os"
"path/filepath"
"strings"
"testing"
"github.com/slackhq/nebula/localapi"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestResolveSocketDiagnosesAnOverlongOverride: the CLI applies the daemon's own rule to
// the override, so it says what the daemon's log says rather than dialing a path the
// kernel would truncate and reporting the generic failure. The rule is sun_path's, which
// a named pipe does not have, so this is a unix test.
func TestResolveSocketDiagnosesAnOverlongOverride(t *testing.T) {
cfg := filepath.Join(t.TempDir(), "config.yml")
long := "/" + strings.Repeat("x", 120)
require.NoError(t, os.WriteFile(cfg, []byte("local_api:\n control:\n socket: "+long+"\n"), 0600))
_, err := resolveSocket("", cfg)
require.Error(t, err)
require.ErrorContains(t, err, "too long for this platform")
assert.ErrorContains(t, err, localapi.CtlSocketKey)
}
// TestResolveSocketRefusesARelativeOverride: the CLI applies the daemon's rule here too,
// so a relative path in the config is reported as the daemon's log reports it rather
// than resolved against wherever nebula ctl happened to be run from.
func TestResolveSocketRefusesARelativeOverride(t *testing.T) {
cfg := filepath.Join(t.TempDir(), "config.yml")
require.NoError(t, os.WriteFile(cfg, []byte("local_api:\n control:\n socket: run/ctl.sock\n"), 0600))
_, err := resolveSocket("", cfg)
require.Error(t, err)
require.ErrorContains(t, err, "must be an absolute path")
assert.ErrorContains(t, err, localapi.CtlSocketKey)
}
+2 -24
View File
@@ -8,9 +8,7 @@ import (
"strings" "strings"
"github.com/slackhq/nebula" "github.com/slackhq/nebula"
"github.com/slackhq/nebula/cmd/internal/ctl"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/localapi"
"github.com/slackhq/nebula/logging" "github.com/slackhq/nebula/logging"
"github.com/slackhq/nebula/util" "github.com/slackhq/nebula/util"
) )
@@ -34,11 +32,6 @@ func init() {
} }
func main() { func main() {
// commands are either `nebula ctl ...` or `nebula -<flag> ...`
if code, done := ctl.Subcommand(os.Args); done {
os.Exit(code)
}
serviceFlag := flag.String("service", "", "Control the system service.") serviceFlag := flag.String("service", "", "Control the system service.")
configPath := flag.String("config", "", "Path to either a file or directory to load configuration from") configPath := flag.String("config", "", "Path to either a file or directory to load configuration from")
configTest := flag.Bool("test", false, "Test the config and print the end result. Non zero exit indicates a faulty config") configTest := flag.Bool("test", false, "Test the config and print the end result. Non zero exit indicates a faulty config")
@@ -47,12 +40,6 @@ func main() {
flag.Parse() flag.Parse()
// A word after the daemon's flags is `ctl` in the other place an operator puts it,
// or a mistyped subcommand, and either way not a daemon to start; see Trailing.
if code, done := ctl.Trailing(flag.Args(), *configPath); done {
os.Exit(code)
}
if *printVersion { if *printVersion {
fmt.Printf("Version: %s\n", Build) fmt.Printf("Version: %s\n", Build)
os.Exit(0) os.Exit(0)
@@ -98,10 +85,9 @@ func main() {
fmt.Printf("failed to apply logging config: %s", err) fmt.Printf("failed to apply logging config: %s", err)
os.Exit(1) os.Exit(1)
} }
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) { c.RegisterReloadCallback(func(c *config.C) {
if err := logging.ApplyConfig(l, c); err != nil { if err := logging.ApplyConfig(l, c); err != nil {
l.Error("Failed to reconfigure logger on reload", "error", err) l.Error("Failed to reconfigure logger on reload", "error", err)
r.Failed()
} }
}) })
@@ -117,17 +103,9 @@ func main() {
os.Exit(1) os.Exit(1)
} }
ctl := localapi.Wire(ctrl.Context(), l, c, ctrl.Commands())
go ctrl.ShutdownBlock() go ctrl.ShutdownBlock()
err := ctrl.Wait() if err := ctrl.Wait(); err != nil {
// os.Exit skips deferred calls, so the socket is removed here, before either exit
// below, rather than left to a deferred Stop that would never run.
ctl.Stop()
if err != nil {
l.Error("Nebula stopped due to fatal error", "error", err) l.Error("Nebula stopped due to fatal error", "error", err)
os.Exit(2) os.Exit(2)
} }
+1 -15
View File
@@ -8,7 +8,6 @@ import (
"github.com/kardianos/service" "github.com/kardianos/service"
"github.com/slackhq/nebula" "github.com/slackhq/nebula"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/localapi"
"github.com/slackhq/nebula/logging" "github.com/slackhq/nebula/logging"
) )
@@ -18,7 +17,6 @@ type program struct {
configPath *string configPath *string
build string build string
control *nebula.Control control *nebula.Control
ctl *localapi.Server
} }
func (p *program) Start(s service.Service) error { func (p *program) Start(s service.Service) error {
@@ -36,10 +34,9 @@ func (p *program) Start(s service.Service) error {
if err := logging.ApplyConfig(l, c); err != nil { if err := logging.ApplyConfig(l, c); err != nil {
return fmt.Errorf("failed to apply logging config: %s", err) return fmt.Errorf("failed to apply logging config: %s", err)
} }
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) { c.RegisterReloadCallback(func(c *config.C) {
if err := logging.ApplyConfig(l, c); err != nil { if err := logging.ApplyConfig(l, c); err != nil {
l.Error("Failed to reconfigure logger on reload", "error", err) l.Error("Failed to reconfigure logger on reload", "error", err)
r.Failed()
} }
}) })
@@ -52,16 +49,10 @@ func (p *program) Start(s service.Service) error {
return err return err
} }
// Now that we're started, stand up the local control socket.
p.ctl = localapi.Wire(p.control.Context(), l, c, p.control.Commands())
// Nebula can stop itself on a fatal packet reader error, make sure to log it if it happens. // Nebula can stop itself on a fatal packet reader error, make sure to log it if it happens.
go func() { go func() {
if err := p.control.Wait(); err != nil { if err := p.control.Wait(); err != nil {
logger.Error(fmt.Sprintf("Nebula stopped due to fatal error: %v", err)) logger.Error(fmt.Sprintf("Nebula stopped due to fatal error: %v", err))
// os.Exit skips deferred calls and races the socket's own context watcher,
// so the socket is removed here, as cmd/nebula does.
p.ctl.Stop()
os.Exit(2) os.Exit(2)
} }
}() }()
@@ -71,11 +62,6 @@ func (p *program) Start(s service.Service) error {
func (p *program) Stop(s service.Service) error { func (p *program) Stop(s service.Service) error {
logger.Info("Nebula service stopping.") logger.Info("Nebula service stopping.")
if p.ctl != nil {
p.ctl.Stop()
}
if p.control == nil { if p.control == nil {
return nil return nil
} }
+2 -24
View File
@@ -8,9 +8,7 @@ import (
"strings" "strings"
"github.com/slackhq/nebula" "github.com/slackhq/nebula"
"github.com/slackhq/nebula/cmd/internal/ctl"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/localapi"
"github.com/slackhq/nebula/logging" "github.com/slackhq/nebula/logging"
"github.com/slackhq/nebula/util" "github.com/slackhq/nebula/util"
) )
@@ -34,11 +32,6 @@ func init() {
} }
func main() { func main() {
// commands are either `nebula ctl ...` or `nebula -<flag> ...`
if code, done := ctl.Subcommand(os.Args); done {
os.Exit(code)
}
configPath := flag.String("config", "", "Path to either a file or directory to load configuration from") configPath := flag.String("config", "", "Path to either a file or directory to load configuration from")
configTest := flag.Bool("test", false, "Test the config and print the end result. Non zero exit indicates a faulty config") configTest := flag.Bool("test", false, "Test the config and print the end result. Non zero exit indicates a faulty config")
printVersion := flag.Bool("version", false, "Print version") printVersion := flag.Bool("version", false, "Print version")
@@ -46,12 +39,6 @@ func main() {
flag.Parse() flag.Parse()
// A word after the daemon's flags is `ctl` in the other place an operator puts it,
// or a mistyped subcommand, and either way not a daemon to start; see Trailing.
if code, done := ctl.Trailing(flag.Args(), *configPath); done {
os.Exit(code)
}
if *printVersion { if *printVersion {
fmt.Printf("Version: %s\n", Build) fmt.Printf("Version: %s\n", Build)
os.Exit(0) os.Exit(0)
@@ -84,10 +71,9 @@ func main() {
fmt.Printf("failed to apply logging config: %s", err) fmt.Printf("failed to apply logging config: %s", err)
os.Exit(1) os.Exit(1)
} }
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) { c.RegisterReloadCallback(func(c *config.C) {
if err := logging.ApplyConfig(l, c); err != nil { if err := logging.ApplyConfig(l, c); err != nil {
l.Error("Failed to reconfigure logger on reload", "error", err) l.Error("Failed to reconfigure logger on reload", "error", err)
r.Failed()
} }
}) })
@@ -103,18 +89,10 @@ func main() {
os.Exit(1) os.Exit(1)
} }
ctl := localapi.Wire(ctrl.Context(), l, c, ctrl.Commands())
go ctrl.ShutdownBlock() go ctrl.ShutdownBlock()
notifyReady(l) notifyReady(l)
err := ctrl.Wait() if err := ctrl.Wait(); err != nil {
// os.Exit skips deferred calls, so the socket is removed here, before either exit
// below, rather than left to a deferred Stop that would never run.
ctl.Stop()
if err != nil {
l.Error("Nebula stopped due to fatal error", "error", err) l.Error("Nebula stopped due to fatal error", "error", err)
os.Exit(2) os.Exit(2)
} }
-624
View File
@@ -1,624 +0,0 @@
package nebula
import (
"encoding/json"
"flag"
"fmt"
"io"
"log/slog"
"maps"
"net/netip"
"sort"
"strings"
"github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/header"
"github.com/slackhq/nebula/logging"
)
// The commands nebula serves. attachCommands builds the one registry every surface
// reads; the ssh debug server is a consumer of it, not its owner, which is why this
// is not ssh.go. See the commands package for what a Command is.
type listHostMapFlags struct {
ByIndex bool
}
type printCertFlags struct {
Raw bool
}
type changeRemoteFlags struct {
Address string
}
type closeTunnelFlags struct {
LocalOnly bool
}
type createTunnelFlags struct {
Address string
}
// lighthouseInfo is one entry of list-lighthouse-addrmap.
type lighthouseInfo struct {
VpnAddr string `json:"vpnAddr"`
Addrs *CacheMap `json:"addrs"`
}
// deviceInfo is the result of device-info.
type deviceInfo struct {
Name string `json:"name"`
Cidr []netip.Prefix `json:"cidr"`
}
// certPEM is the result of print-cert -raw. The PEM form is a different thing to emit,
// not a different formatting of the certificate, so it is its own shape.
type certPEM struct {
PEM string `json:"pem"`
}
// Result shapes for print-relays. Field names are the JSON keys; changing them would
// change the output.
type relayFor struct {
Error string `json:",omitempty"`
Type string
State string
PeerAddr netip.Addr
LocalIndex uint32
RemoteIndex uint32
RelayedThrough []netip.Addr
}
type relayOutput struct {
NebulaAddr netip.Addr
RelayForAddrs []relayFor
}
type relaysOutput struct {
Relays []*relayOutput
}
type versionInfo struct {
Version string `json:"version"`
}
type logLevelInfo struct {
Level string `json:"level"`
}
type logFormatInfo struct {
Format string `json:"format"`
}
// defineListHostMapFlags declares the flags list-hostmap and list-pending-hostmap share.
func defineListHostMapFlags(fl *flag.FlagSet, s *listHostMapFlags) {
fl.BoolVar(&s.ByIndex, "by-index", false, "gets all hosts in the hostmap from the index table")
}
func attachCommands(l *slog.Logger, c *config.C, reg *commands.Registry, f *Interface) {
reg.Register(commands.Flagged(&commands.Command{
Name: commands.ListHostmap,
ShortDescription: "List all known previously connected hosts",
Text: commands.Typed(textListHostMap),
}, defineListHostMapFlags, func(s *listHostMapFlags, a []string) (commands.Result, error) {
return runListHostMap(f.hostMap, s)
}))
reg.Register(commands.Flagged(&commands.Command{
Name: commands.ListPendingHostmap,
ShortDescription: "List all handshaking hosts",
Text: commands.Typed(textListHostMap),
}, defineListHostMapFlags, func(s *listHostMapFlags, a []string) (commands.Result, error) {
return runListHostMap(f.handshakeManager, s)
}))
reg.Register(&commands.Command{
Name: commands.ListLighthouseAddrmap,
ShortDescription: "List all lighthouse map entries",
Run: func(fs any, a []string) (commands.Result, error) {
return runListLighthouseMap(f.lightHouse)
},
Text: commands.Typed(textListLighthouseMap),
})
reg.Register(&commands.Command{
Name: commands.Reload,
ShortDescription: "Reloads configuration from disk, same as sending HUP to the process",
// The line goes out before the reload: a reload that restarts the ssh server ends
// the session that asked, and this is what it sees.
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Did("Reloading config"), nil
},
Then: func() commands.Result {
return runReload(l, c)
},
})
reg.Register(&commands.Command{
Name: commands.LogLevel,
ShortDescription: "Gets or sets the current log level",
Run: func(fs any, a []string) (commands.Result, error) {
return runLogLevel(l, a)
},
Text: commands.Typed(textLogLevel),
})
reg.Register(&commands.Command{
Name: commands.LogFormat,
ShortDescription: "Gets or sets the current log format",
Run: func(fs any, a []string) (commands.Result, error) {
return runLogFormat(l, a)
},
Text: commands.Typed(textLogFormat),
})
reg.Register(&commands.Command{
Name: commands.Version,
ShortDescription: "Prints the currently running version of nebula",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Value(versionInfo{Version: f.version}), nil
},
Text: commands.Typed(textVersion),
})
reg.Register(&commands.Command{
Name: commands.DeviceInfo,
ShortDescription: "Prints information about the network device.",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Value(runDeviceInfo(f)), nil
},
Text: commands.Typed(textDeviceInfo),
})
reg.Register(commands.Flagged(&commands.Command{
Name: commands.PrintCert,
ShortDescription: "Prints the current certificate being used or the certificate for the provided vpn addr",
Text: textPrintCert,
}, func(fl *flag.FlagSet, s *printCertFlags) {
fl.BoolVar(&s.Raw, "raw", false, "prints the PEM encoded certificate; with -json, an object holding it")
}, func(s *printCertFlags, a []string) (commands.Result, error) {
return runPrintCert(f, s, a)
}))
reg.Register(&commands.Command{
Name: commands.PrintTunnel,
ShortDescription: "Prints json details about a tunnel for the provided vpn addr",
Run: func(fs any, a []string) (commands.Result, error) {
return runPrintTunnel(f, a)
},
// No Text on purpose: nothing human readable exists yet, and shipping an interim
// form would invite scripts against it. ssh defaults these to JSON; ctl asks.
JSONOnly: true,
})
reg.Register(&commands.Command{
Name: commands.PrintRelays,
ShortDescription: "Prints json details about all relay info",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Value(runPrintRelays(f)), nil
},
// No Text on purpose; see print-tunnel.
JSONOnly: true,
})
reg.Register(commands.Flagged(&commands.Command{
Name: commands.ChangeRemote,
ShortDescription: "Changes the remote address used in the tunnel for the provided vpn addr",
}, func(fl *flag.FlagSet, s *changeRemoteFlags) {
fl.StringVar(&s.Address, "address", "", "The new remote address, ip:port")
}, func(s *changeRemoteFlags, a []string) (commands.Result, error) {
return runChangeRemote(f, s, a)
}))
reg.Register(commands.Flagged(&commands.Command{
Name: commands.CloseTunnel,
ShortDescription: "Closes a tunnel for the provided vpn addr",
}, func(fl *flag.FlagSet, s *closeTunnelFlags) {
fl.BoolVar(&s.LocalOnly, "local-only", false, "Disables notifying the remote that the tunnel is shutting down")
}, func(s *closeTunnelFlags, a []string) (commands.Result, error) {
return runCloseTunnel(f, s, a)
}))
reg.Register(commands.Flagged(&commands.Command{
Name: commands.CreateTunnel,
ShortDescription: "Creates a tunnel for the provided vpn address",
Help: "The lighthouses will be queried for real addresses but you can provide one as well.",
}, func(fl *flag.FlagSet, s *createTunnelFlags) {
fl.StringVar(&s.Address, "address", "", "Optionally provide a real remote address, ip:port ")
}, func(s *createTunnelFlags, a []string) (commands.Result, error) {
return runCreateTunnel(f, s, a)
}))
reg.Register(&commands.Command{
Name: commands.QueryLighthouse,
ShortDescription: "Query the lighthouses for the provided vpn address",
Help: "This command is asynchronous. Only currently known udp addresses will be printed.",
Run: func(fs any, a []string) (commands.Result, error) {
return runQueryLighthouse(f, a)
},
// No Text on purpose; see print-tunnel.
JSONOnly: true,
})
}
// runListHostMap computes the hostmap listing. Presentation is Render's job.
func runListHostMap(hl controlHostLister, fs *listHostMapFlags) (commands.Result, error) {
var hm []ControlHostInfo
if fs.ByIndex {
hm = listHostMapIndexes(hl)
} else {
hm = listHostMapHosts(hl)
}
sort.Slice(hm, func(i, j int) bool {
return hm[i].VpnAddrs[0].Compare(hm[j].VpnAddrs[0]) < 0
})
return commands.Value(hm), nil
}
// textListHostMap is the human rendering, shared by every surface.
func textListHostMap(hm []ControlHostInfo, w io.Writer) error {
for _, h := range hm {
if _, err := fmt.Fprintf(w, "%s: %s\n", h.VpnAddrs, h.RemoteAddrs); err != nil {
return err
}
}
return nil
}
func runListLighthouseMap(lightHouse *LightHouse) (commands.Result, error) {
lightHouse.RLock()
addrMap := make([]lighthouseInfo, len(lightHouse.addrMap))
x := 0
for k, v := range lightHouse.addrMap {
addrMap[x] = lighthouseInfo{
VpnAddr: k.String(),
Addrs: v.CopyCache(),
}
x++
}
lightHouse.RUnlock()
sort.Slice(addrMap, func(i, j int) bool {
return strings.Compare(addrMap[i].VpnAddr, addrMap[j].VpnAddr) < 0
})
return commands.Value(addrMap), nil
}
func textListLighthouseMap(addrMap []lighthouseInfo, w io.Writer) error {
for _, e := range addrMap {
b, err := json.Marshal(e.Addrs)
if err != nil {
return err
}
if _, err := fmt.Fprintf(w, "%s: %s\n", e.VpnAddr, string(b)); err != nil {
return err
}
}
return nil
}
func textVersion(vi versionInfo, w io.Writer) error {
_, err := fmt.Fprintf(w, "%s\n", vi.Version)
return err
}
// vpnAddrOperand parses the vpn address a command takes as its first operand. noun is
// how the command's messages name it, "vpn address" or "vpn addr", which the output
// golden pins per command. With no operand, or one that does not parse, ok is false and
// res is the refusal to return.
func vpnAddrOperand(a []string, noun string) (vpnAddr netip.Addr, res commands.Result, ok bool) {
if len(a) == 0 {
return netip.Addr{}, commands.Failed("No vpn address was provided"), false
}
vpnAddr, err := netip.ParseAddr(a[0])
if err != nil {
return netip.Addr{}, commands.Failed("The provided %s could not be parsed: %s", noun, a[0]), false
}
return vpnAddr, commands.Result{}, true
}
// tunnelOperand is vpnAddrOperand followed by the lookup of the tunnel it names.
func tunnelOperand(ifce *Interface, a []string, noun string) (*HostInfo, commands.Result, bool) {
vpnAddr, res, ok := vpnAddrOperand(a, noun)
if !ok {
return nil, res, false
}
hostInfo := ifce.hostMap.QueryVpnAddr(vpnAddr)
if hostInfo == nil {
return nil, commands.Failed("Could not find tunnel for %s: %v", noun, a[0]), false
}
return hostInfo, commands.Result{}, true
}
func runQueryLighthouse(ifce *Interface, a []string) (commands.Result, error) {
vpnAddr, res, ok := vpnAddrOperand(a, "vpn address")
if !ok {
return res, nil
}
var cm *CacheMap
rl := ifce.lightHouse.Query(vpnAddr)
if rl != nil {
cm = rl.CopyCache()
}
return commands.Value(cm), nil
}
func runCloseTunnel(ifce *Interface, flags *closeTunnelFlags, a []string) (commands.Result, error) {
hostInfo, res, ok := tunnelOperand(ifce, a, "vpn address")
if !ok {
return res, nil
}
if !flags.LocalOnly {
ifce.send(
header.CloseTunnel,
0,
hostInfo.ConnectionState,
hostInfo,
[]byte{},
make([]byte, 12, 12),
make([]byte, mtu),
)
}
ifce.closeTunnel(hostInfo)
return commands.Did("Closed"), nil
}
func runCreateTunnel(ifce *Interface, flags *createTunnelFlags, a []string) (commands.Result, error) {
vpnAddr, res, ok := vpnAddrOperand(a, "vpn address")
if !ok {
return res, nil
}
if hostInfo := ifce.hostMap.QueryVpnAddr(vpnAddr); hostInfo != nil {
return commands.Failed("Tunnel already exists"), nil
}
if hostInfo := ifce.handshakeManager.QueryVpnAddr(vpnAddr); hostInfo != nil {
return commands.Failed("Tunnel already handshaking"), nil
}
var addr netip.AddrPort
if flags.Address != "" {
var err error
addr, err = netip.ParseAddrPort(flags.Address)
if err != nil {
return commands.Failed("Address could not be parsed"), nil
}
}
// Set the remote from the callback, which StartHandshake runs under the handshake manager lock
// before the hostinfo reaches the handshake loop or the static-host trigger, so the address is
// in the RemoteList before the first handshake is sent.
ifce.handshakeManager.StartHandshake(vpnAddr, func(hh *HandshakeHostInfo) {
if addr.IsValid() {
hh.hostinfo.SetRemote(addr)
}
})
return commands.Did("Created"), nil
}
func runChangeRemote(ifce *Interface, flags *changeRemoteFlags, a []string) (commands.Result, error) {
if len(a) == 0 {
return commands.Failed("No vpn address was provided"), nil
}
if flags.Address == "" {
return commands.Failed("No address was provided"), nil
}
addr, err := netip.ParseAddrPort(flags.Address)
if err != nil {
return commands.Failed("Address could not be parsed"), nil
}
hostInfo, res, ok := tunnelOperand(ifce, a, "vpn address")
if !ok {
return res, nil
}
hostInfo.SetRemote(addr)
return commands.Did("Changed"), nil
}
func runLogLevel(l *slog.Logger, a []string) (commands.Result, error) {
ctrl, ok := l.Handler().(interface {
GetLevel() slog.Level
SetLevel(slog.Level)
})
if !ok {
return commands.Failed("Log level is not reconfigurable on this logger"), nil
}
if len(a) > 0 {
level, err := logging.ParseLevel(strings.ToLower(a[0]))
if err != nil {
return commands.Failed("Unknown log level %s. Possible log levels: trace, debug, info, warn, error", a), nil
}
ctrl.SetLevel(level)
}
return commands.Value(logLevelInfo{Level: logging.LevelName(ctrl.GetLevel())}), nil
}
func textLogLevel(li logLevelInfo, w io.Writer) error {
_, err := fmt.Fprintf(w, "Log level is: %s\n", li.Level)
return err
}
func runLogFormat(l *slog.Logger, a []string) (commands.Result, error) {
ctrl, ok := l.Handler().(interface {
GetFormat() string
SetFormat(string) error
})
if !ok {
return commands.Failed("Log format is not reconfigurable on this logger"), nil
}
if len(a) > 0 {
if err := ctrl.SetFormat(strings.ToLower(a[0])); err != nil {
return commands.Failed("Unable to set log format: %s", err), nil
}
}
return commands.Value(logFormatInfo{Format: ctrl.GetFormat()}), nil
}
func textLogFormat(fi logFormatInfo, w io.Writer) error {
_, err := fmt.Fprintf(w, "Log format is: %s\n", fi.Format)
return err
}
func runPrintCert(ifce *Interface, args *printCertFlags, a []string) (commands.Result, error) {
c := ifce.pki.getCertState().GetDefaultCertificate()
if len(a) > 0 {
hostInfo, res, ok := tunnelOperand(ifce, a, "vpn addr")
if !ok {
return res, nil
}
// A tunnel still handshaking, or one being torn down, has no certificate.
cc := hostInfo.GetCert()
if cc == nil {
return commands.Failed("Could not find tunnel for vpn addr: %v", a[0]), nil
}
c = cc.Certificate
}
if args.Raw {
b, err := c.MarshalPEM()
if err != nil {
return commands.Failed("Unable to marshal the certificate: %s", err), nil
}
return commands.Value(certPEM{PEM: string(b)}), nil
}
return commands.Value(c), nil
}
// textPrintCert takes the value untyped: print-cert returns a certificate or, with -raw,
// its PEM, and those are two things to emit rather than two renderings of one.
func textPrintCert(v any, w io.Writer) error {
switch c := v.(type) {
case certPEM:
_, err := io.WriteString(w, c.PEM)
return err
case cert.Certificate:
_, err := io.WriteString(w, c.String()+"\n")
return err
default:
return fmt.Errorf("expected a certificate, got %T", v)
}
}
func runPrintRelays(ifce *Interface) any {
ifce.hostMap.RLock()
relays := maps.Clone(ifce.hostMap.Relays)
ifce.hostMap.RUnlock()
co := relaysOutput{}
for k, v := range relays {
ro := relayOutput{NebulaAddr: v.vpnAddrs[0]}
co.Relays = append(co.Relays, &ro)
relayHI := ifce.hostMap.QueryVpnAddr(v.vpnAddrs[0])
if relayHI == nil {
ro.RelayForAddrs = append(ro.RelayForAddrs, relayFor{Error: "could not find hostinfo"})
continue
}
for _, vpnAddr := range relayHI.relayState.CopyRelayForIps() {
rf := relayFor{}
r, ok := relayHI.relayState.GetRelayForByAddr(vpnAddr)
if ok {
t := ""
switch r.Type {
case ForwardingType:
t = "forwarding"
case TerminalType:
t = "terminal"
default:
t = "unknown"
}
st := ""
switch r.State {
case Requested:
st = "requested"
case Established:
st = "established"
default:
st = "unknown"
}
rf.LocalIndex = r.LocalIndex
rf.RemoteIndex = r.RemoteIndex
rf.PeerAddr = r.PeerAddr
rf.Type = t
rf.State = st
if rf.LocalIndex != k {
rf.Error = fmt.Sprintf("hostmap LocalIndex '%v' does not match RelayState LocalIndex", k)
}
}
relayedHI := ifce.hostMap.QueryVpnAddr(vpnAddr)
if relayedHI != nil {
rf.RelayedThrough = append(rf.RelayedThrough, relayedHI.relayState.CopyRelayIps()...)
}
ro.RelayForAddrs = append(ro.RelayForAddrs, rf)
}
}
return co
}
func runPrintTunnel(ifce *Interface, a []string) (commands.Result, error) {
hostInfo, res, ok := tunnelOperand(ifce, a, "vpn addr")
if !ok {
return res, nil
}
return commands.Value(copyHostInfo(hostInfo, ifce.hostMap.GetPreferredRanges())), nil
}
func runDeviceInfo(ifce *Interface) deviceInfo {
d := deviceInfo{
Name: ifce.inside.Name(),
Cidr: make([]netip.Prefix, len(ifce.inside.Networks())),
}
copy(d.Cidr, ifce.inside.Networks())
return d
}
func textDeviceInfo(d deviceInfo, w io.Writer) error {
_, err := fmt.Fprintf(w, "name=%v cidr=%v\n", d.Name, d.Cidr)
return err
}
// runReload reloads the config and reports how it went. A reload that failed is a
// refusal, exit 1 over ssh; one that merely needs a restart is not.
func runReload(l *slog.Logger, c *config.C) commands.Result {
res := c.ReloadConfig()
res.Log(l)
if res.Err != nil || res.Failed {
return commands.Failed("%s", res.String())
}
return commands.Did("%s", res.String())
}
-300
View File
@@ -1,300 +0,0 @@
package commands
import (
"errors"
"flag"
"fmt"
"io"
"strings"
)
// CommandFlags is called before help or command execution to parse command line flags.
// It should return a flag.FlagSet instance and a pointer to the struct that will contain parsed flags
//
// Every call must build a fresh FlagSet and a fresh struct. Nothing may be shared between
// calls: a command runs on more than one surface and on more than one goroutine.
type CommandFlags func() (*flag.FlagSet, any)
// RunFunc computes a command's Result.
//
// fs is the pointer returned by Command.Flags, if there was one; -h, -help, -json and
// -pretty are reserved and handled before Run is called. a is the operands: whatever
// was not consumed as a flag. The Result is rendered by whichever surface asked: Value
// for a command that reports, Did or Failed for one that acts, Nothing for one with
// nothing to say. An error is for a failure of the command itself; a bad argument is
// Failed, so the caller hears why in its own format.
type RunFunc func(fs any, a []string) (Result, error)
// TextFunc renders a Result's Value for a person. See Typed for declaring one against
// the type the command actually returns.
type TextFunc func(v any, w io.Writer) error
type Command struct {
Name Name
ShortDescription string
Help string
Flags CommandFlags
// Run produces the command's Result, leaving presentation to Render.
Run RunFunc
// Text renders a Result's Value for a person. Both surfaces call it, so a command
// reads the same over ssh as it does through the control socket. It never sees an
// outcome: Render prints those itself, the same way for every command. Leaving it
// nil declares that no human rendering of the Value exists yet, which JSONOnly says
// up front; Render then refuses text rather than inventing an interim form for
// people to script against.
Text TextFunc
// JSONOnly declares that the command's Value has no human rendering yet, so a surface
// whose default is text asks for JSON instead, and the schema says so to a client. It
// is a statement that no such form exists, not a fallback. A refusal is still
// rendered as a sentence, since that always has a human form.
JSONOnly bool
// Then, when set, is the second half of a command that acts: Run says what is about
// to happen and Then does it and says how it went, with its Result rendered after
// Run's and deciding the outcome. A Run that refused is the outcome, and Then does
// not run. It exists for a command whose action can cut off
// its own reply: reload restarts the ssh server when the sshd keys changed, which
// ends the session that asked, so Run's line has to be on the wire before the reload
// begins. A surface that answers in one piece, the control socket, runs Then before
// it answers and carries both.
Then func() Result
}
// Failure is the error a surface gets for a refusal the person has already been shown:
// a Run that returned Failed, a command that is not there, a line that would not parse.
// The message it rendered is the whole story, so the surface reports the failure, exit
// status 1 over ssh, and does not log it.
type Failure struct {
Msg string
}
func (f *Failure) Error() string {
return f.Msg
}
// commandFlags returns the FlagSet a command is driven through, on any surface: the
// command's own flags, if it declares any, with the two presentation flags added. own
// reports whether the command declared any of its own, which is what tells a caller
// whether to say a parameter is unknown or that the command takes none at all.
//
// The presentation flags are defined here rather than by each command, because how a
// result is displayed is not a property of the command that produced it. This is also
// the one definition of them: the ssh line parses into it, nebula ctl parses its line
// against the schema derived from it, and help renders from that schema. They are
// command line flags only: a socket request chooses its rendering with the envelope's
// format field, and Invoke refuses them as parameters.
func commandFlags(c *Command) (fl *flag.FlagSet, fs any, own bool) {
if c.Flags != nil {
fl, fs = c.Flags()
}
own = fl != nil
if fl == nil {
fl = flag.NewFlagSet(string(c.Name), flag.ContinueOnError)
}
addPresentationFlags(fl)
// What Parse prints after a bad flag, and for -h: the command's usage as help renders
// it, so a line that went wrong is answered with the same text as `help <command>`.
fl.Usage = func() { _ = WriteCommandHelp(schemaOf(c, fl), fl.Output()) }
return fl, fs, own
}
// addPresentationFlags defines the two presentation flags on fl.
func addPresentationFlags(fl *flag.FlagSet) {
fl.Bool(FlagJSON, false, "outputs as json")
fl.Bool(FlagPretty, false, "pretty prints json, assumes -json")
}
// Flagged fills in c's Flags and Run as a pair: define declares the flags on a fresh F,
// and run is handed that F. The struct the flags fill is the struct Run reads by
// construction, so no Run has to assert the type it was given or decide what to do
// when the assertion fails.
func Flagged[F any](c *Command, define func(*flag.FlagSet, *F), run func(*F, []string) (Result, error)) *Command {
c.Flags = func() (*flag.FlagSet, any) {
fl := flag.NewFlagSet("", flag.ContinueOnError)
s := new(F)
define(fl, s)
return fl, s
}
c.Run = func(fs any, a []string) (Result, error) {
s, ok := fs.(*F)
if !ok {
return Result{}, fmt.Errorf("%s was handed %T for its flags, not %T", c.Name, fs, s)
}
return run(s, a)
}
return c
}
// isPresentationFlag reports whether name is one of the two flags commandFlags adds.
func isPresentationFlag(name string) bool {
return name == FlagJSON || name == FlagPretty
}
// execLine runs an ssh command line against c, reporting problems to w. The line is
// parsed into the command's own FlagSet and Run reads it directly, so a flag.Value whose
// String does not re-parse, flag.Func above all, arrives as the line said. Invoke fills
// the same FlagSet through Set.
func (r *Registry) execLine(c *Command, args []string, w Output) error {
fl, fs, _ := commandFlags(c)
// Parse reports a bad flag to this writer, with the command's usage after it. -h,
// -help and --help take the same path with nothing wrong: the usage is the whole
// answer, and Parse says so with ErrHelp. A -h that is a flag's value or follows --
// is read as the operand it is, since ParseArgs is the one reading of the line.
fl.SetOutput(w.Err)
operands, err := ParseArgs(fl, args)
if err != nil {
if errors.Is(err, flag.ErrHelp) {
return nil
}
return err
}
format := presentationFormat(fl)
// An error from Run is something the command could not do rather than a refusal it
// chose, but the person at the prompt still hears it; the surface logs it too.
res, err := c.Run(fs, operands)
if err != nil {
_, _ = fmt.Fprintln(w.Err, err.Error())
return err
}
// Run's reply goes out first, and then the action, if the command has one and Run
// did not refuse, whose Result is the outcome. In JSON only the outcome travels:
// the promise is for a person to read, and a program expects one document.
if c.Then == nil || !res.OK || r.formatFor(c, res, format) == FormatText {
if err := r.renderLine(c, res, format, w); err != nil {
return err
}
}
if c.Then != nil && res.OK {
res = c.Then()
if err := r.renderLine(c, res, format, w); err != nil {
return err
}
}
if !res.OK {
return &Failure{Msg: res.Reason}
}
return nil
}
// formatFor is the format a Result is rendered in. The line wins when it asked for one;
// otherwise the surface that owns the registry decides, from the command and its
// Result, so each surface can keep its own default.
func (r *Registry) formatFor(c *Command, res Result, format Format) Format {
if format == FormatUnset {
return r.defaultFormatFor(c, res)
}
return format
}
// renderLine writes one Result of a command line to w; see formatFor. An empty Result
// writes nothing.
func (r *Registry) renderLine(c *Command, res Result, format Format, w Output) error {
if res.Empty() {
return nil
}
if err := Render(c, res, r.formatFor(c, res, format), w.Out); err != nil {
if errors.Is(err, ErrNoTextRendering) {
_, _ = fmt.Fprintln(w.Err, err.Error())
}
return err
}
return nil
}
// ParseArgs parses args with fl and returns the operands.
//
// It differs from fl.Parse in two ways, and both surfaces parse through here so a line
// means the same typed at an ssh prompt or to nebula ctl.
//
// Flags and operands may come in any order, which is the one way it differs from
// fl.Parse. Parse stops at the first operand, so `print-tunnel 10.0.0.1 -json` would
// reach the command with "-json" as a second operand and the flag silently dropped. Here
// an operand ends nothing: parsing resumes after it, and only "--" ends the flags, with
// everything after it an operand. What counts as a flag is Parse's rule: every argument
// that starts with "-", a negative number included, so one that is meant as an operand
// goes after "--".
func ParseArgs(fl *flag.FlagSet, args []string) ([]string, error) {
operands := make([]string, 0, len(args))
for len(args) > 0 {
// Walk the flags the way Parse will, to the first argument it would not read as a
// flag: an operand or the terminator. Everything before it is Parse's.
end := len(args)
for i := 0; i < len(args); {
a := args[i]
if a == "--" {
if err := fl.Parse(args[:i]); err != nil {
return nil, err
}
return append(operands, args[i+1:]...), nil
}
if len(a) < 2 || a[0] != '-' {
end = i
break
}
f, hasValue := LookupArg(fl, a)
if f == nil {
// An undefined flag: leave it for Parse to report.
i++
continue
}
if hasValue || IsBoolFlag(f) {
i++
} else {
// The flag's value is the next argument, whatever it looks like.
i += 2
}
}
if err := fl.Parse(args[:end]); err != nil {
return nil, err
}
if end == len(args) {
break
}
operands = append(operands, args[end])
args = args[end+1:]
}
return operands, nil
}
// LookupArg reads arg the way Parse does: one or two leading dashes, the flag's name, and
// an optional "=value". f is the flag of fl by that name, nil when fl has none, and
// hasValue reports that the value came with the argument rather than in the one after it.
// It is the one reading of an argument's shape: ParseArgs walks a line with it, and
// nebula ctl lifts its own flags out of a line with it, so the two cannot disagree about
// which word is a flag's value.
func LookupArg(fl *flag.FlagSet, arg string) (f *flag.Flag, hasValue bool) {
name := strings.TrimPrefix(strings.TrimPrefix(arg, "-"), "-")
if eq := strings.IndexByte(name, '='); eq >= 0 {
name, hasValue = name[:eq], true
}
return fl.Lookup(name), hasValue
}
// IsBoolFlag reports whether f is satisfied by its presence alone and never reads the
// argument after it as its value. The flag package marks these with an unexported
// interface, which is the only way to tell.
func IsBoolFlag(f *flag.Flag) bool {
bf, ok := f.Value.(interface{ IsBoolFlag() bool })
return ok && bf.IsBoolFlag()
}
-101
View File
@@ -1,101 +0,0 @@
package commands
import (
"bytes"
"flag"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestParseArgsAllowsFlagsAfterOperands pins the one way ParseArgs departs from fl.Parse,
// that a flag may follow an operand, and that it departs in no other: a negative number
// is a flag, as Parse reads it, and goes after "--" when it is meant as an operand.
func TestParseArgsAllowsFlagsAfterOperands(t *testing.T) {
cases := map[string]struct {
args []string
wantOps []string
wantN string
wantV bool
wantJSON bool
wantErr string
}{
"flag after operand": {args: []string{"x", "-json"}, wantOps: []string{"x"}, wantJSON: true},
"flags around": {args: []string{"-v", "x", "-n", "1", "y"}, wantOps: []string{"x", "y"}, wantV: true, wantN: "1"},
"value after operand": {args: []string{"x", "-n", "-1"}, wantOps: []string{"x"}, wantN: "-1"},
"equals value": {args: []string{"-n=-1", "x"}, wantOps: []string{"x"}, wantN: "-1"},
"terminator after": {args: []string{"x", "--", "-json", "-1"}, wantOps: []string{"x", "-json", "-1"}},
"terminator": {args: []string{"--", "-json"}, wantOps: []string{"-json"}},
"defined numeric": {args: []string{"-2", "x"}, wantOps: []string{"x"}, wantV: true},
"negative number": {args: []string{"-1"}, wantErr: "flag provided but not defined: -1"},
"bad flag after": {args: []string{"x", "-nope"}, wantErr: "flag provided but not defined: -nope"},
"undefined flag": {args: []string{"-nope", "x"}, wantErr: "flag provided but not defined: -nope"},
"missing value": {args: []string{"-n"}, wantErr: "flag needs an argument: -n"},
}
for name, tc := range cases {
t.Run(name, func(t *testing.T) {
fl := flag.NewFlagSet("t", flag.ContinueOnError)
fl.SetOutput(&bytes.Buffer{})
n := fl.String("n", "", "")
v := fl.Bool("v", false, "")
two := fl.Bool("2", false, "a flag that happens to be a number")
asJSON := fl.Bool(FlagJSON, false, "")
ops, err := ParseArgs(fl, tc.args)
if tc.wantErr != "" {
require.ErrorContains(t, err, tc.wantErr)
return
}
require.NoError(t, err)
assert.Equal(t, tc.wantOps, ops)
assert.Equal(t, tc.wantN, *n)
assert.Equal(t, tc.wantV, *v || *two)
assert.Equal(t, tc.wantJSON, *asJSON)
})
}
}
// TestDispatchFlagAfterOperand: the same through a registered command with no Flags of
// its own, which only has the presentation flags to find.
func TestDispatchFlagAfterOperand(t *testing.T) {
out := &bytes.Buffer{}
require.NoError(t, testRegistry().DispatchArgs([]string{"served", "x", "-json"}, MergedOutput(out)))
assert.Equal(t, "\"read x\"\n", out.String())
}
// TestLookupArg pins the reading of an argument's shape that ParseArgs and nebula ctl
// share.
func TestLookupArg(t *testing.T) {
fl := flag.NewFlagSet("t", flag.ContinueOnError)
fl.String("n", "", "")
fl.Bool("v", false, "")
for _, tc := range []struct {
arg string
name string
hasValue bool
}{
{"-n", "n", false},
{"--n", "n", false},
{"-n=1", "n", true},
{"--n=", "n", true},
{"-v", "v", false},
{"-nope", "", false},
{"-1", "", false},
} {
f, hasValue := LookupArg(fl, tc.arg)
if tc.name == "" {
assert.Nil(t, f, tc.arg)
} else {
require.NotNil(t, f, tc.arg)
assert.Equal(t, tc.name, f.Name, tc.arg)
}
assert.Equal(t, tc.hasValue, hasValue, tc.arg)
}
assert.True(t, IsBoolFlag(fl.Lookup("v")))
assert.False(t, IsBoolFlag(fl.Lookup("n")))
}
-28
View File
@@ -1,28 +0,0 @@
package commands
// Name identifies a command. Every command nebula serves is a constant here, so a
// surface, a client and a test spell it the same way and are told at compile time when
// they do not. It is a string rather than an integer so a test can register a throwaway
// command without adding it here, and so the wire carries the name as the string it has
// always been.
type Name string
const (
ChangeRemote Name = "change-remote"
CloseTunnel Name = "close-tunnel"
CreateTunnel Name = "create-tunnel"
DeviceInfo Name = "device-info"
Help Name = "help"
ListHostmap Name = "list-hostmap"
ListLighthouseAddrmap Name = "list-lighthouse-addrmap"
ListPendingHostmap Name = "list-pending-hostmap"
LogFormat Name = "log-format"
LogLevel Name = "log-level"
Logout Name = "logout"
PrintCert Name = "print-cert"
PrintRelays Name = "print-relays"
PrintTunnel Name = "print-tunnel"
QueryLighthouse Name = "query-lighthouse"
Reload Name = "reload"
Version Name = "version"
)
-428
View File
@@ -1,428 +0,0 @@
package commands
import (
"cmp"
"flag"
"fmt"
"io"
"maps"
"slices"
"strings"
"sync"
"github.com/anmitsu/go-shlex"
)
// Registry holds a set of commands.
//
// It is safe for concurrent use. Commands may be registered while surfaces are serving
// from the registry: the ssh server clones it per session and the control socket filters
// it per request, and either may overlap a Register.
type Registry struct {
mu sync.RWMutex
commands map[Name]*Command
// full is set on a filtered registry and points at the registry it was filtered
// from, so a command that's filtered can be reported as such.
full *Registry
// reason explains, in the terms an operator cares about, why this registry withholds
// a command they can name. It is supplied by whoever built the filter, because why a
// surface declines a command is that surface's business, not the registry's.
reason string
// DefaultFormat decides how a result is rendered when the command line named no
// format. It sees the command and its Result, and applies to values and refusals
// alike: the registry does not special-case either. The renderer is shared between
// surfaces; what each asks for by default is not, so the surface that owns this
// registry sets it. Nil means text.
DefaultFormat func(c *Command, r Result) Format
}
// defaultFormatFor resolves the surface's default for c's Result.
func (r *Registry) defaultFormatFor(c *Command, res Result) Format {
if r.DefaultFormat == nil {
return FormatText
}
return r.DefaultFormat(c, res)
}
// New returns an empty Registry with only `help` available.
func New() *Registry {
r := &Registry{commands: map[Name]*Command{}}
r.registerHelp()
return r
}
func (r *Registry) registerHelp() {
r.Register(&Command{
Name: Help,
ShortDescription: "prints available commands or help <command> for specific usage info",
Run: func(a any, args []string) (Result, error) {
if len(args) == 0 {
return Value(r.Schema()), nil
}
if c := r.Get(Name(args[0])); c != nil {
return Value(r.schemaFor(c)), nil
}
if err := r.unavailable(Name(args[0])); err != nil {
return Failed("%s", err), nil
}
return Failed("Command not available %s", args[0]), nil
},
Text: renderHelp,
})
}
// renderHelp is help's human rendering: the command list, or one command's usage.
func renderHelp(v any, w io.Writer) error {
switch h := v.(type) {
case []CommandSchema:
return writeCommandList(h, w)
case CommandSchema:
return WriteCommandHelp(h, w)
default:
return fmt.Errorf("expected a command schema, got %T", v)
}
}
func writeCommandList(cmds []CommandSchema, w io.Writer) error {
if _, err := io.WriteString(w, "Available commands:\n"); err != nil {
return err
}
// In the order the schema came, which is the display order; see Registry.All.
lines := make([]string, 0, len(cmds))
for _, c := range cmds {
lines = append(lines, fmt.Sprintf("%s - %s", c.Name, c.Short))
}
_, err := io.WriteString(w, strings.Join(lines, "\n")+"\n\n")
return err
}
// WriteCommandHelp renders one command's usage. It is the single rendering of a
// CommandSchema for a person: the daemon uses it for `help <command>` on both surfaces,
// and a client that already holds the schema uses it for `-h`, so the two cannot drift.
func WriteCommandHelp(c CommandSchema, w io.Writer) error {
if _, err := fmt.Fprintf(w, "%s - %s\n", c.Name, c.Short); err != nil {
return err
}
if c.Help != "" {
if _, err := fmt.Fprintf(w, " %s\n", c.Help); err != nil {
return err
}
}
// Same layout as flag.PrintDefaults, including its habit of leaving out a default
// that is the zero value.
for _, f := range c.Flags {
kind := ""
if !f.Bool {
kind = " string"
}
def := ""
if f.Default != "" && f.Default != "false" {
def = fmt.Sprintf(" (default %s)", f.Default)
}
if _, err := fmt.Fprintf(w, " -%s%s\n \t%s%s\n", f.Name, kind, f.Usage, def); err != nil {
return err
}
}
return nil
}
// Register adds a command that can be run by a user.
//
// It panics if the command declares a reserved flag. That is a programming error, and
// surfacing it here, at startup, is better than at the moment an operator first types
// the command.
func (r *Registry) Register(c *Command) {
if c.Run == nil {
panic(fmt.Sprintf("commands: %s has no Run", c.Name))
}
if c.Flags != nil {
if fl, _ := c.Flags(); fl != nil {
for _, name := range reservedFlags {
if fl.Lookup(name) != nil {
panic(fmt.Sprintf("commands: %s declares reserved flag -%s", c.Name, name))
}
}
}
}
r.insert(c)
}
// insert adds c without the reserved flag check, for commands that have already passed
// it on another registry.
func (r *Registry) insert(c *Command) {
r.mu.Lock()
defer r.mu.Unlock()
r.commands[c.Name] = c
}
// Get returns the command registered under name, or nil if there is no such command.
func (r *Registry) Get(name Name) *Command {
r.mu.RLock()
defer r.mu.RUnlock()
return r.commands[name]
}
// Match returns the names of every command starting with prefix, sorted, as strings for
// a prompt to complete and print.
func (r *Registry) Match(prefix string) []string {
cmds := make([]string, 0)
r.mu.RLock()
for n := range r.commands {
if strings.HasPrefix(string(n), prefix) {
cmds = append(cmds, string(n))
}
}
r.mu.RUnlock()
slices.Sort(cmds)
return cmds
}
// All returns every registered command sorted by name, which is the order help lists
// them and the schema describes them.
func (r *Registry) All() []*Command {
r.mu.RLock()
defer r.mu.RUnlock()
cmds := slices.Collect(maps.Values(r.commands))
slices.SortFunc(cmds, func(a, b *Command) int { return cmp.Compare(a.Name, b.Name) })
return cmds
}
// Clone returns a copy holding the same commands.
//
// help is re-registered on the copy rather than carried over, because the source's help
// is bound to the source and would list its commands, not the copy's, once a caller
// registers something on the copy.
func (r *Registry) Clone() *Registry {
r.mu.RLock()
out := &Registry{
commands: maps.Clone(r.commands),
full: r.full,
reason: r.reason,
DefaultFormat: r.DefaultFormat,
}
r.mu.RUnlock()
out.registerHelp()
return out
}
// Filter returns a filtered copy of the registry.
//
// Keep should return true for commands that should be included in the copy.
//
// reason says, to an operator who names a command the copy left out, why this surface
// does not serve it; "" leaves the registry to say only that it does not.
func (r *Registry) Filter(keep func(*Command) bool, reason string) *Registry {
out := &Registry{commands: map[Name]*Command{}, full: r, reason: reason, DefaultFormat: r.DefaultFormat}
for _, cmd := range r.All() {
if keep(cmd) {
out.insert(cmd)
}
}
// Bind help to the filtered set so it only advertises what this surface serves.
out.registerHelp()
return out
}
// Dispatch parses a command line and runs it, reporting problems to w. The error is for
// the caller to log, not to show; see RunFunc and Failure.
func (r *Registry) Dispatch(line string, w Output) error {
// A line shlex cannot split, an unbalanced quote. Reported on Err so the output
// stream stays parseable.
args, err := shlex.Split(line, true)
if err != nil {
msg := "could not parse the command line: " + err.Error()
if _, err := fmt.Fprintln(w.Err, msg); err != nil {
return err
}
return &Failure{Msg: msg}
}
return r.DispatchArgs(args, w)
}
// DispatchArgs runs an already tokenized command line. The error is as for Dispatch.
//
// An empty line and an unknown command both end up running `help`, the first with the
// arguments help itself would take; `<command> -h` is answered by the command's own
// FlagSet, with the same rendering, in execLine.
func (r *Registry) DispatchArgs(args []string, w Output) error {
if len(args) == 0 {
args = []string{"help"}
}
c := r.Get(Name(args[0]))
if c == nil {
if err := r.unavailable(Name(args[0])); err != nil {
// Saying why, rather than "did not understand", keeps the rule legible and
// points at a surface that will actually run the command.
_, _ = fmt.Fprintln(w.Err, err)
return err
}
// Only the first word was looked up, so only it is named.
msg := fmt.Sprintf("did not understand: %s", args[0])
if _, err := fmt.Fprintln(w.Out, msg); err != nil {
return err
}
if err := r.DispatchArgs([]string{"help"}, w); err != nil {
return err
}
return &Failure{Msg: msg}
}
return r.execLine(c, args[1:], w)
}
// unavailable explains why name cannot be run here when this registry was filtered from
// one that holds it: the command exists, and this surface withholds it, for the reason
// whoever built the filter gave. It is nil when name is simply unknown, so the two cases
// read differently to an operator.
func (r *Registry) unavailable(name Name) error {
if r.full == nil {
return nil
}
c := r.full.Get(name)
if c == nil {
return nil
}
reason := r.reason
if reason == "" {
reason = "this surface does not serve it"
}
return fmt.Errorf("%s is not available on this surface: %s", c.Name, reason)
}
// Resolve returns the command registered under name, or says why there is none: the
// filter's reason for one this surface withholds, or unknown for a name that does not
// exist.
func (r *Registry) Resolve(name Name) (*Command, error) {
if c := r.Get(name); c != nil {
return c, nil
}
if err := r.unavailable(name); err != nil {
return nil, err
}
return nil, fmt.Errorf("unknown command %q", name)
}
// setParams sets each named parameter on fl, in the order of names, and stops at the
// first that is unknown or does not parse. own says whether the command declared flags
// of its own, which decides between "takes no parameters" and "has no parameter".
//
// The presentation flags are on fl, since commandFlags put them there, but they are not
// parameters: a request has one place to choose its rendering, the envelope's format
// field, and a client that sends them here is told where that is rather than having a
// second, weaker way to say the same thing.
func setParams(fl *flag.FlagSet, name string, names []string, params map[string]string, own bool) error {
for _, k := range names {
if isPresentationFlag(k) {
return fmt.Errorf("%s is a command line flag, not a parameter; choose the rendering with the request's format field", k)
}
f := fl.Lookup(k)
if f == nil {
if !own {
return fmt.Errorf("%s takes no parameters", name)
}
return fmt.Errorf("%s has no parameter %q", name, k)
}
v := params[k]
if err := fl.Set(k, v); err != nil {
if IsBoolFlag(f) {
// The flag package's own words for this are "parse error".
return fmt.Errorf("%s: expected a boolean, got %q", k, v)
}
return fmt.Errorf("%s: %w", k, err)
}
}
return nil
}
// Schema describes every command in the registry, so a client can offer them without
// having been built against this daemon.
func (r *Registry) Schema() []CommandSchema {
all := r.All()
out := make([]CommandSchema, 0, len(all))
for _, c := range all {
out = append(out, r.schemaFor(c))
}
return out
}
// schemaFor describes one command. It is the one derivation, for the whole schema and
// for `help <command>` alike.
func (r *Registry) schemaFor(c *Command) CommandSchema {
fl, _, _ := commandFlags(c)
return schemaOf(c, fl)
}
// schemaOf describes c as driven through fl, the FlagSet commandFlags built for it. The
// presentation flags are included because a command line client parses its line
// against this and help renders from it; they are listed after the command's own,
// which is where readers expect them. Invoke refuses them as parameters.
func schemaOf(c *Command, fl *flag.FlagSet) CommandSchema {
cs := CommandSchema{Name: c.Name, Short: c.ShortDescription, Help: c.Help}
var own, shared []FlagSchema
fl.VisitAll(func(f *flag.Flag) {
fs := FlagSchema{
Name: f.Name,
Bool: IsBoolFlag(f),
Default: f.DefValue,
Usage: f.Usage,
}
if isPresentationFlag(f.Name) {
shared = append(shared, fs)
return
}
own = append(own, fs)
})
cs.Flags = append(own, shared...)
return cs
}
// CommandSchema describes one command well enough for a client to offer it.
type CommandSchema struct {
Name Name `json:"name"`
Short string `json:"short,omitempty"`
Help string `json:"help,omitempty"`
Flags []FlagSchema `json:"flags,omitempty"`
}
// FlagSchema describes one of a command's named parameters.
type FlagSchema struct {
Name string `json:"name"`
// Bool reports a parameter that takes no value, which is what tells a command line
// client whether `-pretty` swallows the word after it.
Bool bool `json:"bool,omitempty"`
Default string `json:"default,omitempty"`
Usage string `json:"usage,omitempty"`
}
-420
View File
@@ -1,420 +0,0 @@
package commands
import (
"bytes"
"errors"
"flag"
"io"
"strings"
"sync"
"testing"
"time"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// notWithheld is a stand in for a surface's own policy, which the registry knows nothing
// of; see Command.Identity for the one the info API applies.
func notWithheld(c *Command) bool { return c.Name != "withheld" }
const testReason = "this surface does not run commands that write"
// textLine is a Text renderer for test commands whose result is one string.
func textLine(v any, w io.Writer) error {
_, err := io.WriteString(w, v.(string)+"\n")
return err
}
func testRegistry() *Registry {
r := New()
r.Register(&Command{
Name: "served",
ShortDescription: "a command this surface serves",
Run: func(fs any, a []string) (Result, error) {
return Value("read " + strings.Join(a, ",")), nil
},
Text: textLine,
})
r.Register(&Command{
Name: "withheld",
ShortDescription: "a command this surface withholds",
Run: func(fs any, a []string) (Result, error) {
return Value("wrote"), nil
},
Text: textLine,
})
return r
}
// TestFilterKeepsWhatKeepSays: the copy holds the commands keep accepted, and help.
func TestFilterKeepsWhatKeepSays(t *testing.T) {
ro := testRegistry().Filter(notWithheld, testReason)
names := make([]string, 0)
for _, c := range ro.All() {
names = append(names, string(c.Name))
}
assert.ElementsMatch(t, []string{"help", "served"}, names)
}
// TestRefusesWithheldByName covers a client that knows a command exists and asks for it
// anyway. The filter lives on the serving side, so the answer must not depend on what
// the client was told was available.
func TestRefusesWithheldByName(t *testing.T) {
out, errOut := &bytes.Buffer{}, &bytes.Buffer{}
w := Output{Out: out, Err: errOut}
err := testRegistry().Filter(notWithheld, testReason).DispatchArgs([]string{"withheld"}, w)
require.Error(t, err)
assert.Contains(t, err.Error(), "withheld is not available on this surface")
assert.Empty(t, out.String(), "a withheld command must produce no output")
assert.Contains(t, errOut.String(), "withheld is not available on this surface")
assert.Contains(t, errOut.String(), testReason,
"the registry should render the explanation the filter supplied, not one of its own")
}
// TestWithheldIsNotReportedAsUnknown keeps the two cases distinct: an operator who
// mistypes and one who asked for a real command should not get the same answer.
func TestWithheldIsNotReportedAsUnknown(t *testing.T) {
out, errOut := &bytes.Buffer{}, &bytes.Buffer{}
err := testRegistry().Filter(notWithheld, testReason).DispatchArgs([]string{"no-such-command"}, Output{Out: out, Err: errOut})
var f *Failure
require.ErrorAs(t, err, &f, "an unknown command is a failure to the surface")
assert.Contains(t, out.String(), "did not understand: no-such-command")
}
// paramRegistry holds one command with a bool flag and a duration flag, and reports what
// the command saw when it ran.
func paramRegistry(t *testing.T) (*Registry, *[]paramsSeen) {
t.Helper()
var seen []paramsSeen
r := New()
r.Register(&Command{
Name: "set",
ShortDescription: "sets things",
Flags: func() (*flag.FlagSet, any) {
fl := flag.NewFlagSet("", flag.ContinueOnError)
s := paramsSeen{}
fl.BoolVar(&s.Verbose, "verbose", false, "say more")
fl.DurationVar(&s.Wait, "wait", 0, "how long")
return fl, &s
},
Run: func(fs any, a []string) (Result, error) {
seen = append(seen, *fs.(*paramsSeen))
return Value("ok"), nil
},
})
return r, &seen
}
type paramsSeen struct {
Verbose bool
Wait time.Duration
}
// TestInvokeReportsParametersInNameOrder: parameters arrive as an unordered object, so
// with two bad ones the error must not depend on map iteration. Name order decides.
func TestInvokeReportsParametersInNameOrder(t *testing.T) {
r, _ := paramRegistry(t)
cases := map[string]struct {
params map[string]string
want string
}{
"two unknown": {map[string]string{"zed": "1", "alpha": "1"}, `set has no parameter "alpha"`},
"unknown before bad bool": {map[string]string{"aaa": "1", "verbose": "maybe"}, `set has no parameter "aaa"`},
"bad bool before unknown": {map[string]string{"verbose": "maybe", "zzz": "1"}, `verbose: expected a boolean, got "maybe"`},
"bad value before bad bool": {map[string]string{"wait": "soon", "verbose": "maybe"}, `verbose: expected a boolean`},
}
for name, tc := range cases {
t.Run(name, func(t *testing.T) {
// Enough runs that a map-order dependence would show.
for i := 0; i < 50; i++ {
_, err := Invoke(Request{Command: r.Get("set"), Params: tc.params})
require.Error(t, err)
assert.ErrorContains(t, err, tc.want)
}
})
}
}
// TestInvokeWithABadParameterSetsNothing: a request with any bad parameter must not run
// the command, and must leave nothing behind for the next request to see. The good
// parameter sorts before the bad one, so a one-pass Invoke would have applied it.
func TestInvokeWithABadParameterSetsNothing(t *testing.T) {
r, seen := paramRegistry(t)
_, err := Invoke(Request{Command: r.Get("set"), Params: map[string]string{"verbose": "true", "wait": "soon"}})
require.ErrorContains(t, err, "wait:")
assert.Empty(t, *seen, "the command must not run with a partial parameter set")
v, err := Invoke(Request{Command: r.Get("set")})
require.NoError(t, err)
assert.Equal(t, "ok", v.Value)
require.Len(t, *seen, 1)
assert.Equal(t, paramsSeen{}, (*seen)[0], "nothing from the failed request carried over")
_, err = Invoke(Request{Command: r.Get("set"), Params: map[string]string{"verbose": "true", "wait": "2s"}})
require.NoError(t, err)
require.Len(t, *seen, 2)
assert.Equal(t, paramsSeen{Verbose: true, Wait: 2 * time.Second}, (*seen)[1])
}
func TestHelpListsOnlyServedCommands(t *testing.T) {
out := &bytes.Buffer{}
require.NoError(t, testRegistry().Filter(notWithheld, testReason).DispatchArgs(nil, MergedOutput(out)))
assert.Contains(t, out.String(), "served")
assert.NotContains(t, out.String(), "withheld")
}
// TestDispatchReportsUnparseableLine pins that a line shlex rejects says so on Err rather
// than coming back silent, so an operator with an unbalanced quote learns the quote is the
// problem instead of assuming the command is broken.
func TestDispatchReportsUnparseableLine(t *testing.T) {
out, errOut := &bytes.Buffer{}, &bytes.Buffer{}
testRegistry().Dispatch(`served "unterminated`, Output{Out: out, Err: errOut})
assert.Empty(t, out.String())
assert.Contains(t, errOut.String(), "could not parse the command line")
}
// TestRegisterRejectsReservedFlags: a command that declares a reserved flag is refused
// at Register rather than when an operator first runs it; see reservedFlags.
func TestRegisterRejectsReservedFlags(t *testing.T) {
for _, name := range []string{"json", "pretty", "h", "help", "config", "socket"} {
t.Run(name, func(t *testing.T) {
r := New()
assert.PanicsWithValue(t, "commands: clash declares reserved flag -"+name, func() {
r.Register(&Command{
Name: "clash",
Flags: func() (*flag.FlagSet, any) {
fl := flag.NewFlagSet("clash", flag.ContinueOnError)
fl.Bool(name, false, "")
return fl, nil
},
Run: func(fs any, a []string) (Result, error) { return Nothing(), nil },
})
})
})
}
}
// TestRegistryIsSafeForConcurrentUse runs the operations the surfaces perform, filtering
// and cloning per request or session, against a stream of registrations. It exists to
// be run under the race detector: the map underneath is not goroutine safe.
func TestRegistryIsSafeForConcurrentUse(t *testing.T) {
r := testRegistry()
var wg sync.WaitGroup
for i := 0; i < 8; i++ {
wg.Add(1)
go func(i int) {
defer wg.Done()
for j := 0; j < 50; j++ {
r.Register(&Command{
Name: Name("late-" + strings.Repeat("a", i) + "-" + strings.Repeat("b", j)),
ShortDescription: "registered after the surfaces started",
Run: func(fs any, a []string) (Result, error) { return Nothing(), nil },
})
}
}(i)
}
for i := 0; i < 8; i++ {
wg.Add(1)
go func() {
defer wg.Done()
for j := 0; j < 50; j++ {
f := r.Filter(notWithheld, testReason)
_ = f.Get("served")
_ = f.Schema()
_ = r.Clone().Match("late-")
}
}()
}
wg.Wait()
f := r.Filter(notWithheld, testReason)
assert.Len(t, f.Match("late-"), 8*50, "every late registration is visible once the writers finish")
}
// TestCloneHelpSeesCloneCommands: the ssh session registers logout on its clone, and
// the source does not learn of it. Help on the clone answers for the clone, and -h on a
// command prints what `help <name>` does, on a clone as on the source: both render
// through WriteCommandHelp, and Clone rebinds help to the copy.
func TestCloneHelpSeesCloneCommands(t *testing.T) {
r := testRegistry()
c := r.Clone()
c.Register(&Command{
Name: Logout,
ShortDescription: "ends the session",
Run: func(fs any, a []string) (Result, error) { return Nothing(), nil },
})
assert.Nil(t, r.Get(Logout), "the source is not affected")
out := &bytes.Buffer{}
require.NoError(t, c.DispatchArgs([]string{"help"}, MergedOutput(out)))
assert.Contains(t, out.String(), "logout")
for _, name := range []string{"logout", "served"} {
flag := &bytes.Buffer{}
c.Dispatch(name+" -h", MergedOutput(flag))
cmd := &bytes.Buffer{}
c.Dispatch("help "+name, MergedOutput(cmd))
assert.Equal(t, cmd.String(), flag.String(), name)
assert.Contains(t, flag.String(), name+" - ", name)
}
}
// TestThenRunsAfterTheReply pins the order Then exists for: Run's line is on the writer
// before the action begins, Then's Result follows it, and Then's outcome is the
// command's.
func TestThenRunsAfterTheReply(t *testing.T) {
out := &bytes.Buffer{}
seen := "never ran"
outcome := Did("acted")
r := New()
r.Register(&Command{
Name: "act",
ShortDescription: "replies, then acts",
Run: func(fs any, a []string) (Result, error) {
return Did("about to act"), nil
},
Then: func() Result {
seen = out.String()
return outcome
},
})
require.NoError(t, r.DispatchArgs([]string{"act"}, MergedOutput(out)))
assert.Equal(t, "about to act\n", seen, "the line was written before the action")
assert.Equal(t, "about to act\nacted\n", out.String())
out.Reset()
outcome = Failed("could not act")
var f *Failure
require.ErrorAs(t, r.DispatchArgs([]string{"act"}, MergedOutput(out)), &f)
assert.Equal(t, "about to act\ncould not act\n", out.String())
// In JSON only the outcome travels: one document for the program that asked.
out.Reset()
require.ErrorAs(t, r.DispatchArgs([]string{"act", "-json"}, MergedOutput(out)), &f)
assert.JSONEq(t, `{"ok":false,"reason":"could not act"}`, out.String())
// A Run that refused is the outcome; the action does not happen.
ran := false
r.Register(&Command{
Name: "refuse",
ShortDescription: "refuses, so never acts",
Run: func(fs any, a []string) (Result, error) {
return Failed("not like that"), nil
},
Then: func() Result {
ran = true
return Did("acted anyway")
},
})
out.Reset()
require.ErrorAs(t, r.DispatchArgs([]string{"refuse"}, MergedOutput(out)), &f)
assert.Equal(t, "not like that\n", out.String())
assert.False(t, ran)
out.Reset()
require.ErrorAs(t, r.DispatchArgs([]string{"refuse", "-json"}, MergedOutput(out)), &f)
assert.JSONEq(t, `{"ok":false,"reason":"not like that"}`, out.String())
assert.False(t, ran)
}
// TestWriteCommandHelpIsTheOneRendering pins the layout both `help <command>` on the
// daemon and `-h` in the client print, so the two cannot drift apart again.
func TestWriteCommandHelpIsTheOneRendering(t *testing.T) {
cs := CommandSchema{
Name: "change-remote",
Short: "changes a remote",
Help: "Give it an address.",
Flags: []FlagSchema{
{Name: "address", Usage: "the new remote", Default: "1.2.3.4:4242"},
{Name: "by-index", Bool: true, Usage: "match by index"},
{Name: "json", Bool: true, Usage: "outputs as json"},
},
}
out := &bytes.Buffer{}
require.NoError(t, WriteCommandHelp(cs, out))
assert.Equal(t, "change-remote - changes a remote\n"+
" Give it an address.\n"+
" -address string\n \tthe new remote (default 1.2.3.4:4242)\n"+
" -by-index\n \tmatch by index\n"+
" -json\n \toutputs as json\n", out.String())
}
// addrFlags is the flag struct of a test command that takes -address.
type addrFlags struct{ Address string }
// TestHelpFlagIsReadWhereParseReadsIt: -h asks for help only where the line's parser
// reads it as a flag. As a flag's value or after -- it reaches the command as the
// operand it is, and help is rendered by the same function `help <command>` uses.
func TestHelpFlagIsReadWhereParseReadsIt(t *testing.T) {
var seen []string
r := New()
r.Register(Flagged(&Command{Name: "x", ShortDescription: "takes an address"},
func(fl *flag.FlagSet, s *addrFlags) { fl.StringVar(&s.Address, "address", "", "an address") },
func(s *addrFlags, a []string) (Result, error) {
seen = append(seen, s.Address+"|"+strings.Join(a, ","))
return Nothing(), nil
}))
for _, line := range []string{"x -h", "x -help", "x --help", "x 10.0.0.1 -h", "x -address=1 -h"} {
out := &bytes.Buffer{}
r.Dispatch(line, MergedOutput(out))
assert.Contains(t, out.String(), "x - takes an address\n", line)
assert.Contains(t, out.String(), "-address string", line)
}
assert.Empty(t, seen, "asking for help must not run the command")
for _, line := range []string{"x -address -h", "x -- -h"} {
out := &bytes.Buffer{}
r.Dispatch(line, MergedOutput(out))
assert.Empty(t, out.String(), line)
}
assert.Equal(t, []string{"-h|", "|-h"}, seen)
}
// TestBadFlagIsAnsweredWithTheCommandsHelp: a flag the command does not have is reported
// on the error stream followed by the command's usage, in the rendering help uses.
func TestBadFlagIsAnsweredWithTheCommandsHelp(t *testing.T) {
out, errOut := &bytes.Buffer{}, &bytes.Buffer{}
r := New()
r.Register(&Command{Name: "x", ShortDescription: "has no flags", Run: func(fs any, a []string) (Result, error) { return Nothing(), nil }})
require.Error(t, r.DispatchArgs([]string{"x", "-nope"}, Output{Out: out, Err: errOut}))
assert.Empty(t, out.String())
assert.Equal(t, "flag provided but not defined: -nope\nx - has no flags\n -json\n \toutputs as json\n -pretty\n \tpretty prints json, assumes -json\n", errOut.String())
}
// TestRunErrorIsShown: an error Run returns reaches the person at the prompt on the
// error stream, as well as the surface that logs it.
func TestRunErrorIsShown(t *testing.T) {
r := New()
r.Register(&Command{
Name: "broken",
ShortDescription: "cannot do what it is asked",
Run: func(fs any, a []string) (Result, error) {
return Result{}, errors.New("no certificate is loaded")
},
})
out, errOut := &bytes.Buffer{}, &bytes.Buffer{}
require.EqualError(t, r.DispatchArgs([]string{"broken"}, Output{Out: out, Err: errOut}), "no certificate is loaded")
assert.Empty(t, out.String())
assert.Equal(t, "no certificate is loaded\n", errOut.String())
}
-179
View File
@@ -1,179 +0,0 @@
package commands
import (
"encoding/json"
"errors"
"fmt"
"io"
)
// ErrNoTextRendering is returned by Render when a command's Value has no human rendering
// and text was asked for anyway. A command declares this by leaving Text nil and saying
// so with JSONOnly; it is not a fallback situation but a statement that no such form
// exists yet, so nobody builds a script around an interim one. Each surface's default
// policy avoids ever asking.
var ErrNoTextRendering = errors.New("no human readable output yet; pass -json or -pretty")
// Format is how a result should be presented.
type Format int
const (
// FormatUnset is a request that expressed no preference. A surface resolves it with
// its own default before rendering; Render itself treats it as FormatText.
FormatUnset Format = iota
// FormatText renders for a person. It is what an operator gets by default on every
// surface with a human at the other end.
FormatText
// FormatJSON renders the result as compact JSON.
FormatJSON
// FormatPrettyJSON renders the result as indented JSON.
FormatPrettyJSON
)
// Flag names for the two presentation flags. They are offered on every command rather
// than declared by each one, because how a result is displayed is not a property of the
// command that produced it.
const (
FlagJSON = "json"
FlagPretty = "pretty"
)
// Flag names nebula ctl takes for itself, anywhere on its line, to say which daemon to
// talk to; see reservedFlags.
const (
FlagConfig = "config"
FlagSocket = "socket"
)
// reservedFlags are handled before a command's Run is called, so a command may not
// declare them itself. -h and -help are intercepted as a request for usage; -json and
// -pretty are added to every command's FlagSet by commandFlags, on every surface, so a
// command that defined either would make that definition panic with "flag redefined"
// on whichever goroutine ran the command; and -config and -socket are nebula ctl's own,
// taken off the line before the client knows which command it names, so a command
// that declared either would never receive it there.
var reservedFlags = []string{"h", "help", FlagJSON, FlagPretty, FlagConfig, FlagSocket}
// Render writes r in the requested format. Every surface calls it, so a hostmap looks
// the same over ssh as it does through the control socket, and a change to either
// rendering happens once.
func Render(c *Command, r Result, f Format, w io.Writer) error {
switch f {
case FormatJSON:
return renderJSON(r.Payload(), w, false)
case FormatPrettyJSON:
return renderJSON(r.Payload(), w, true)
default:
// An outcome reads the same for every command: its reason, as a sentence. Only a
// Value needs the command's own rendering, so Text never sees anything else.
if r.Value == nil {
_, err := io.WriteString(w, r.Reason+"\n")
return err
}
if c != nil && c.Text != nil {
return c.Text(r.Value, w)
}
name := "this command"
if c != nil {
name = string(c.Name)
}
return fmt.Errorf("%s has %w", name, ErrNoTextRendering)
}
}
func renderJSON(v any, w io.Writer, indent bool) error {
enc := json.NewEncoder(w)
if indent {
enc.SetIndent("", " ")
}
return enc.Encode(v)
}
// Result is what every Run returns: whether the command worked, a sentence saying so,
// and, for a command that reports rather than acts, the value it produced.
//
// Value and Reason are alternatives. A command that reports returns Value(v), and the
// surfaces render v, as JSON or through the command's Text. A command that acts, or any
// command refusing its arguments, returns Did or Failed, and the surfaces render the
// Outcome: its Reason is the sentence these commands have always printed. Nothing is for
// a command with nothing to say.
//
// The envelope itself never reaches a caller. Payload is what JSON carries, and Render
// hands Text only the Value, so a command's renderer sees the one type it returns on
// success and is never asked to recognise an Outcome.
type Result struct {
Outcome
Value any
}
// Outcome is the part of a Result that says whether the command worked and, in words a
// caller can show someone, why. It is what a program receives from a command that acts
// rather than reports: closing a tunnel, changing a remote, reloading config.
//
// One shape serves all of them. Designing a payload per outcome would be work without a
// reader. Reason is set either way, so a success says what it did rather than leaving
// the caller to infer it from a bare true.
type Outcome struct {
OK bool `json:"ok"`
Reason string `json:"reason"`
}
// Value is the Result of a command that reports: v is what it has to show, and it must
// marshal to JSON.
func Value(v any) Result {
return Result{Outcome: Outcome{OK: true}, Value: v}
}
// Failed is a convenience for the common "it did not work, here is why" case.
func Failed(format string, args ...any) Result {
return Result{Outcome: Outcome{Reason: fmt.Sprintf(format, args...)}}
}
// Did is a convenience for the common "it worked, here is what happened" case.
func Did(format string, args ...any) Result {
return Result{Outcome: Outcome{OK: true, Reason: fmt.Sprintf(format, args...)}}
}
// Nothing is the Result of a command with nothing to say: it did its work, and the
// surface sends an empty, successful reply without rendering anything.
func Nothing() Result {
return Result{Outcome: Outcome{OK: true}}
}
// Empty reports a Result with nothing to render: no Value and no Reason. The zero Result
// counts, so a Run that returns Result{} says nothing rather than failing silently.
func (r Result) Empty() bool {
return r.Value == nil && r.Reason == ""
}
// Payload is what a program receives: the Value when there is one, otherwise the
// Outcome. No surface sends the envelope itself, so print-tunnel -json emits the tunnel
// and close-tunnel -json emits {ok, reason}.
func (r Result) Payload() any {
if r.Value != nil {
return r.Value
}
return r.Outcome
}
// Typed adapts a renderer for one result type to the TextFunc a Command declares. The
// value Render passes is the command's own, see Result, so all that is left is to name
// its type once here instead of asserting it at the top of every renderer. A renderer
// paired with a command that returns something else fails here, saying what arrived.
func Typed[T any](render func(T, io.Writer) error) TextFunc {
return func(v any, w io.Writer) error {
t, ok := v.(T)
if !ok {
return fmt.Errorf("expected %T, got %T", t, v)
}
return render(t, w)
}
}
-105
View File
@@ -1,105 +0,0 @@
package commands
import (
"bytes"
"io"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestRenderPrintsAnOutcomeWithoutText: a command that acts declares no Text. Its
// outcome renders as the reason, the same for every command, and as {ok, reason} in
// JSON, which is the shape these commands have always emitted.
func TestRenderPrintsAnOutcomeWithoutText(t *testing.T) {
c := &Command{Name: "act"}
out := &bytes.Buffer{}
require.NoError(t, Render(c, Did("Closed"), FormatText, out))
assert.Equal(t, "Closed\n", out.String())
out.Reset()
require.NoError(t, Render(c, Failed("No vpn address was provided"), FormatText, out))
assert.Equal(t, "No vpn address was provided\n", out.String())
out.Reset()
require.NoError(t, Render(c, Failed("nope"), FormatJSON, out))
assert.JSONEq(t, `{"ok":false,"reason":"nope"}`, out.String())
}
// TestRenderHandsTextOnlyTheValue pins the guarantee Result exists for: a command's
// renderer sees the type it returns on success and never the envelope or an outcome, so
// a report command that refuses its arguments needs nothing in its Text to cope.
func TestRenderHandsTextOnlyTheValue(t *testing.T) {
var got any
c := &Command{
Name: "report",
Text: func(v any, w io.Writer) error {
got = v
_, err := io.WriteString(w, "rendered\n")
return err
},
}
out := &bytes.Buffer{}
require.NoError(t, Render(c, Value("data"), FormatText, out))
assert.Equal(t, "data", got)
assert.Equal(t, "rendered\n", out.String())
got, out = nil, &bytes.Buffer{}
require.NoError(t, Render(c, Failed("refused"), FormatText, out))
assert.Nil(t, got, "Text must not be called for an outcome")
assert.Equal(t, "refused\n", out.String())
out.Reset()
require.NoError(t, Render(c, Value("data"), FormatJSON, out))
assert.JSONEq(t, `"data"`, out.String(), "JSON carries the value, never the envelope")
}
// TestRenderRefusesTextForAValueWithoutText: leaving Text nil is a statement that no
// human form of the value exists yet, and asking for one is an error that says so.
func TestRenderRefusesTextForAValueWithoutText(t *testing.T) {
out := &bytes.Buffer{}
err := Render(&Command{Name: "print-tunnel"}, Value(map[string]int{"index": 7}), FormatText, out)
require.ErrorIs(t, err, ErrNoTextRendering)
require.ErrorContains(t, err, "print-tunnel has no human readable output yet")
assert.Empty(t, out.String())
}
// TestTypedNamesTheTypeOnce: Typed replaces the assertion at the top of every renderer.
// The right type reaches the renderer; the wrong one is reported, naming both.
func TestTypedNamesTheTypeOnce(t *testing.T) {
type info struct{ Name string }
render := Typed(func(i info, w io.Writer) error {
_, err := io.WriteString(w, i.Name+"\n")
return err
})
out := &bytes.Buffer{}
require.NoError(t, render(info{Name: "tun0"}, out))
assert.Equal(t, "tun0\n", out.String())
err := render("not an info", out)
require.Error(t, err)
assert.ErrorContains(t, err, "expected commands.info, got string")
}
// TestResultShapes pins what each constructor means to a surface: Empty is the one that
// renders nothing, Payload is what a program receives, and OK is set wherever a reason
// or value says the command worked.
func TestResultShapes(t *testing.T) {
assert.True(t, Nothing().Empty())
assert.True(t, Result{}.Empty(), "the zero Result says nothing rather than failing silently")
assert.False(t, Did("done").Empty())
assert.False(t, Failed("no").Empty())
assert.False(t, Value("x").Empty())
assert.True(t, Value("x").OK)
assert.True(t, Did("done").OK)
assert.False(t, Failed("no").OK)
assert.Equal(t, "x", Value("x").Payload())
assert.Equal(t, Outcome{OK: true, Reason: "done"}, Did("done").Payload())
assert.Equal(t, Outcome{Reason: "no"}, Failed("no").Payload())
}
-119
View File
@@ -1,119 +0,0 @@
package commands
import (
"flag"
"maps"
"slices"
)
// Request is one command to run, as every surface hands it to Invoke: the command, its
// options by name in the form their flags parse, its operands, and the rendering the
// line asked for, if any.
//
// nebula ctl builds one with ParseLine and it travels as a JSON object to the control
// socket. An ssh line never becomes one: it is parsed straight into the command's own
// FlagSet; see execLine.
type Request struct {
// Command is the command to run, resolved by the surface at its edge. See
// Registry.Resolve for the one that tells an unknown name from a withheld one.
Command *Command
// Params carries the options the line set, by flag name, each in the form its flag
// parses from. An option left alone is absent, so the command's default applies.
Params map[string]string
// Args carries the operands: whatever the line did not consume as a flag.
Args []string
// Format is the rendering the line asked for with -json or -pretty, or FormatUnset
// when it expressed no preference. It is ParseLine's reading of the line for the
// client that sends the Request; Invoke does not read it, since the surface that
// runs a Request renders with the format it agreed with its caller.
Format Format
}
// ParseLine reads a FlagSet that ParseArgs has filled, and the operands it returned,
// into a Request to send over the wire. Command is left for the caller to set.
//
// Every flag the line set other than the presentation flags travels by name, as its
// flag's String, which is what the daemon's Set parses. That round trip is exact for
// the bool and string flags a client builds from the schema; it is the client's to make,
// and the ssh path does not make it, since it runs the FlagSet it parsed.
func ParseLine(fl *flag.FlagSet, operands []string) Request {
req := Request{Args: operands, Format: presentationFormat(fl)}
fl.Visit(func(f *flag.Flag) {
if isPresentationFlag(f.Name) {
return
}
if req.Params == nil {
req.Params = map[string]string{}
}
req.Params[f.Name] = f.Value.String()
})
return req
}
// presentationFormat is the one reading of -json and -pretty from a parsed line, so a
// line means the same on every surface: only a flag set to true asks for JSON, since one
// set to false expresses no preference, and -pretty implies -json. FormatUnset means the
// line expressed no preference and the surface's own default applies.
func presentationFormat(fl *flag.FlagSet) Format {
var asJSON, pretty bool
fl.Visit(func(f *flag.Flag) {
switch f.Name {
case FlagJSON:
asJSON = flagIsTrue(f)
case FlagPretty:
pretty = flagIsTrue(f)
}
})
switch {
case pretty:
return FormatPrettyJSON
case asJSON:
return FormatJSON
default:
return FormatUnset
}
}
// flagIsTrue reports whether a bool flag was set to true. The flag package's bool value
// is a flag.Getter whose Get returns the bool, so this never falls back to parsing text.
func flagIsTrue(f *flag.Flag) bool {
g, ok := f.Value.(flag.Getter)
if !ok {
return false
}
b, _ := g.Get().(bool)
return b
}
// Invoke runs one Request and returns the command's Result, unrendered.
//
// The parameters are applied by flag name, in name order: a request from the wire is a
// JSON object with no order of its own, and which of two bad parameters is reported must
// not change from one request to the next. An unknown parameter is an error rather than
// something quietly ignored, because a caller that misspells one deserves to hear about
// it instead of getting a default. A bad parameter means Run is never called, and the
// FlagSet is fresh for this call, so nothing set before it is seen by anyone.
//
// Nothing here writes anywhere and nothing here renders. What a surface does with the
// Result, marshal its Payload or run the command's own text rendering through Render,
// is the surface's decision, and keeping that out of here is what lets a caller tell a
// refusal from a success after the text has been produced. The presentation flags are
// not parameters: a Request says how it wants rendering in Format, and sending them as
// parameters is refused rather than read as a preference.
func Invoke(req Request) (Result, error) {
c := req.Command
names := slices.Sorted(maps.Keys(req.Params))
fl, fs, own := commandFlags(c)
if err := setParams(fl, string(c.Name), names, req.Params, own); err != nil {
return Result{}, err
}
return c.Run(fs, req.Args)
}
-115
View File
@@ -1,115 +0,0 @@
package commands
import (
"flag"
"strings"
"testing"
"time"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func lineFlags() *flag.FlagSet {
fl := flag.NewFlagSet("set", flag.ContinueOnError)
fl.Bool("verbose", false, "")
fl.Duration("wait", 0, "")
fl.String("label", "none", "")
fl.Bool(FlagJSON, false, "")
fl.Bool(FlagPretty, false, "")
return fl
}
// TestParseLineReadsThePresentationFlagsOnce: presentationFormat is the one reading of
// -json and -pretty, for every surface, and ParseLine carries its answer. A flag set to
// false expresses no preference, and neither flag ever travels as a parameter.
func TestParseLineReadsThePresentationFlagsOnce(t *testing.T) {
for _, tc := range []struct {
args []string
want Format
}{
{nil, FormatUnset},
{[]string{"-json=false"}, FormatUnset},
{[]string{"-pretty=false"}, FormatUnset},
{[]string{"-json"}, FormatJSON},
{[]string{"-pretty"}, FormatPrettyJSON},
{[]string{"-pretty", "-json=false"}, FormatPrettyJSON},
} {
fl := lineFlags()
operands, err := ParseArgs(fl, tc.args)
require.NoError(t, err)
req := ParseLine(fl, operands)
assert.Equal(t, tc.want, req.Format, "%v", tc.args)
assert.Nil(t, req.Params, "%v: presentation flags are not parameters", tc.args)
}
}
// TestParseLineCarriesWhatWasSetInParseableForm: only the flags the line set travel,
// each as the string its flag parses from, so the daemon's Invoke can apply them to a
// fresh FlagSet and get the same values back. Operands come through untouched, those
// after "--" included.
func TestParseLineCarriesWhatWasSetInParseableForm(t *testing.T) {
fl := lineFlags()
operands, err := ParseArgs(fl, []string{"-verbose", "-wait", "2s", "x", "--", "-1"})
require.NoError(t, err)
req := ParseLine(fl, operands)
assert.Equal(t, map[string]string{"verbose": "true", "wait": "2s"}, req.Params, "label was left alone and must be absent")
assert.Equal(t, []string{"x", "-1"}, req.Args)
assert.Equal(t, FormatUnset, req.Format)
// The round trip a client makes: what ParseLine produced is what Invoke applies.
r, seen := paramRegistry(t)
req.Command = r.Get("set")
_, err = Invoke(req)
require.NoError(t, err)
require.Len(t, *seen, 1)
assert.Equal(t, paramsSeen{Verbose: true, Wait: 2 * time.Second}, (*seen)[0])
}
// TestResolveTellsUnknownFromWithheld: the two cases read differently to an operator,
// and Resolve is the edge that decides which one a name from the wire is.
func TestResolveTellsUnknownFromWithheld(t *testing.T) {
r := testRegistry().Filter(notWithheld, testReason)
c, err := r.Resolve("served")
require.NoError(t, err)
assert.Equal(t, Name("served"), c.Name)
_, err = r.Resolve("withheld")
require.ErrorContains(t, err, "withheld is not available on this surface")
_, err = r.Resolve("no-such-command")
require.ErrorContains(t, err, `unknown command "no-such-command"`)
}
// TestLineRunsTheFlagSetItParsed: the ssh path hands Run the FlagSet the line was parsed
// into, with no trip through a Request. A flag.Func has no String worth re-parsing, so a
// round trip would hand Run an empty value and no error; the line's value must arrive.
func TestLineRunsTheFlagSetItParsed(t *testing.T) {
var seen []string
r := New()
r.Register(&Command{
Name: "mode",
Flags: func() (*flag.FlagSet, any) {
fl := flag.NewFlagSet("", flag.ContinueOnError)
s := new(string)
fl.Func("mode", "the mode", func(v string) error {
*s = "set:" + v
return nil
})
return fl, s
},
Run: func(fs any, a []string) (Result, error) {
seen = append(seen, *fs.(*string))
return Value("ok"), nil
},
})
var out, errw strings.Builder
r.Dispatch("mode -json -mode fast", Output{Out: &out, Err: &errw})
require.Equal(t, []string{"set:fast"}, seen)
assert.Empty(t, errw.String())
assert.Equal(t, "\"ok\"\n", out.String(), "-json was read from the same FlagSet")
}
-23
View File
@@ -1,23 +0,0 @@
package commands
import "io"
// Output is where a command line surface's text goes.
//
// Out and Err are distinct so a command's primary output can stay machine readable on
// a surface that separates the two. The ssh surface points both at the same channel and
// sees a single merged stream. The local
// control socket does not use one at all: it invokes by named parameter and carries the
// returned value, so there is no stream for a command to write to.
type Output struct {
// Out receives what a command produces.
Out io.Writer
// Err receives usage text, refusals and errors.
Err io.Writer
}
// MergedOutput sends output and errors to the same writer.
func MergedOutput(w io.Writer) Output {
return Output{Out: w, Err: w}
}
-111
View File
@@ -1,111 +0,0 @@
package nebula
import (
"bytes"
"encoding/json"
"io"
"net/netip"
"testing"
"time"
"github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/cert_test"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/logging"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestLogFormatRefusalIsShown: a format the logger does not know is a refusal the user
// sees, rather than a bare error, which over ssh was silence.
func TestLogFormatRefusalIsShown(t *testing.T) {
l := logging.NewLogger(io.Discard)
res, err := runLogFormat(l, []string{"bogus"})
require.NoError(t, err)
assert.False(t, res.OK)
assert.Equal(t, "Unable to set log format: unknown log format `bogus`. possible formats: [text json]", res.Reason)
res, err = runLogFormat(l, []string{"JSON"})
require.NoError(t, err)
out := &bytes.Buffer{}
require.NoError(t, commands.Typed(textLogFormat)(res.Value, out))
assert.Equal(t, "Log format is: json\n", out.String())
}
// TestPrintRelaysReportsErrorsAsText: the Error field was an error value, which encodes
// as {} and lost the message. It is the message now, and absent when there is none.
func TestPrintRelaysReportsErrorsAsText(t *testing.T) {
l := test.NewLogger()
hm := newHostMap(l)
relay := netip.MustParseAddr("10.0.0.1")
peer := netip.MustParseAddr("10.0.0.2")
// A relay host that is in Relays but not in Hosts, so its hostinfo cannot be found.
hm.Relays[1] = &HostInfo{vpnAddrs: []netip.Addr{relay}}
ifce := &Interface{hostMap: hm}
marshal := func() string {
b, err := json.Marshal(runPrintRelays(ifce))
require.NoError(t, err)
return string(b)
}
assert.JSONEq(t, `{"Relays":[{"NebulaAddr":"10.0.0.1","RelayForAddrs":[{"Error":"could not find hostinfo","Type":"","State":"","PeerAddr":"","LocalIndex":0,"RemoteIndex":0,"RelayedThrough":null}]}]}`, marshal())
// The same relay host, found, relaying for a peer whose index matches: no Error key.
hi := hm.Relays[1]
r := &Relay{Type: ForwardingType, State: Established, PeerAddr: peer, LocalIndex: 1, RemoteIndex: 2}
hi.relayState = RelayState{
relayForByAddr: map[netip.Addr]*Relay{peer: r},
relayForByIdx: map[uint32]*Relay{r.LocalIndex: r},
}
hm.Hosts[relay] = hi
assert.JSONEq(t, `{"Relays":[{"NebulaAddr":"10.0.0.1","RelayForAddrs":[{"Type":"forwarding","State":"established","PeerAddr":"10.0.0.2","LocalIndex":1,"RemoteIndex":2,"RelayedThrough":null}]}]}`, marshal())
// And an index mismatch names it.
hm.Relays[7] = hi
delete(hm.Relays, 1)
assert.Contains(t, marshal(), `"Error":"hostmap LocalIndex '7' does not match RelayState LocalIndex"`)
}
// tunnelFixture is an Interface with our own certificate loaded, one peer behind an
// established tunnel and one still handshaking, which has no certificate yet.
func tunnelFixture(t *testing.T) *Interface {
t.Helper()
l := test.NewLogger()
me := netip.MustParseAddr("10.0.0.1")
peer := netip.MustParsePrefix("10.0.0.2/24")
ca, _, caKey, _ := cert_test.NewTestCaCert(cert.Version2, cert.Curve_CURVE25519, time.Time{}, time.Time{}, nil, nil, nil)
peerCert, _, _, _ := cert_test.NewTestCert(cert.Version2, cert.Curve_CURVE25519, ca, caKey, "peer", time.Time{}, time.Time{},
[]netip.Prefix{peer}, nil, []string{"web", "db"})
hm := newHostMap(l)
hm.Hosts[peer.Addr()] = &HostInfo{
vpnAddrs: []netip.Addr{peer.Addr()},
ConnectionState: &ConnectionState{peerCert: &cert.CachedCertificate{Certificate: peerCert}},
}
hm.Hosts[netip.MustParseAddr("10.0.0.3")] = &HostInfo{
vpnAddrs: []netip.Addr{netip.MustParseAddr("10.0.0.3")},
ConnectionState: &ConnectionState{},
}
return &Interface{pki: newTestPKI(t, "me", []netip.Addr{me}), hostMap: hm}
}
// TestPrintCertOfATunnelWithoutACertificate: a tunnel still handshaking, or one being
// torn down, has no certificate, and print-cert refuses rather than panicking.
func TestPrintCertOfATunnelWithoutACertificate(t *testing.T) {
f := tunnelFixture(t)
r, err := runPrintCert(f, &printCertFlags{}, []string{"10.0.0.3"})
require.NoError(t, err)
assert.False(t, r.OK)
assert.Equal(t, "Could not find tunnel for vpn addr: 10.0.0.3", r.Reason)
r, err = runPrintCert(f, &printCertFlags{}, []string{"10.0.0.2"})
require.NoError(t, err)
assert.True(t, r.OK, "a tunnel with a certificate is printed")
}
+23 -65
View File
@@ -26,17 +26,9 @@ type C struct {
files []string files []string
Settings map[string]any Settings map[string]any
oldSettings map[string]any oldSettings map[string]any
callbacks []func(*C, *Reload) callbacks []func(*C)
l *slog.Logger l *slog.Logger
reloadLock sync.Mutex reloadLock sync.Mutex
// Set by the first reload that needs a restart and never cleared. Guarded by reloadLock
needsRestart bool
// callbackLock guards callbacks. Registration is not confined to startup: CatchHUP is
// running from inside Main, and anything wired up after Main returns registers while
// a HUP could already be reloading, so append and iteration have to be serialized.
callbackLock sync.Mutex
} }
func NewC(l *slog.Logger) *C { func NewC(l *slog.Logger) *C {
@@ -80,45 +72,16 @@ func (c *C) LoadString(raw string) error {
// RegisterReloadCallback stores a function to be called when a config reload is triggered. The functions registered // RegisterReloadCallback stores a function to be called when a config reload is triggered. The functions registered
// here should decide if they need to make a change to the current process before making the change. HasChanged can be // here should decide if they need to make a change to the current process before making the change. HasChanged can be
// used to help decide if a change is necessary. // used to help decide if a change is necessary.
// These functions should return quickly or spawn their own go routine if they will take a while. // These functions should return quickly or spawn their own go routine if they will take a while
// Use r to report anything that failed or needs a restart to apply. func (c *C) RegisterReloadCallback(f func(*C)) {
func (c *C) RegisterReloadCallback(f func(c *C, r *Reload)) {
c.callbackLock.Lock()
defer c.callbackLock.Unlock()
c.callbacks = append(c.callbacks, f) c.callbacks = append(c.callbacks, f)
} }
// reloadCallbacks returns a snapshot of the registered callbacks. Reloads iterate the
// snapshot rather than the live slice, so a registration from another goroutine cannot
// race the loop; a callback registered while a reload is in progress runs from the next
// reload on. callbackLock is not held during the calls themselves, so a callback that
// registers another callback does not deadlock.
func (c *C) reloadCallbacks() []func(*C, *Reload) {
c.callbackLock.Lock()
defer c.callbackLock.Unlock()
return append(make([]func(*C, *Reload), 0, len(c.callbacks)), c.callbacks...)
}
// InitialLoad returns true if this is the first load of the config, and ReloadConfig has not been called yet. // InitialLoad returns true if this is the first load of the config, and ReloadConfig has not been called yet.
func (c *C) InitialLoad() bool { func (c *C) InitialLoad() bool {
return c.oldSettings == nil return c.oldSettings == nil
} }
// Render returns keys and their values as yaml, for a subsystem to remember what it applied. Empty means they couldn't
// be rendered and should never match.
func (c *C) Render(keys ...string) string {
m := make(map[string]any, len(keys))
for _, k := range keys {
m[k] = c.Get(k)
}
b, err := yaml.Marshal(m)
if err != nil {
c.l.Debug("Failed to render config", "keys", keys, "error", err)
return ""
}
return string(b)
}
// HasChanged checks if the underlying structure of the provided key has changed after a config reload. The value of // HasChanged checks if the underlying structure of the provided key has changed after a config reload. The value of
// k in both the old and new settings will be serialized, the result of the string comparison is returned. // k in both the old and new settings will be serialized, the result of the string comparison is returned.
// If k is an empty string the entire config is tested. // If k is an empty string the entire config is tested.
@@ -181,55 +144,50 @@ func (c *C) CatchHUP(ctx context.Context) {
return return
case <-ch: case <-ch:
c.l.Info("Caught HUP, reloading config") c.l.Info("Caught HUP, reloading config")
c.ReloadConfig().Log(c.l) c.ReloadConfig()
} }
} }
}() }()
} }
// ReloadConfig reloads the config from the path provided to Load and runs every registered callback. A config that func (c *C) ReloadConfig() {
// fails to load is logged here, like any other failure, and the result carries the error for the summary.
func (c *C) ReloadConfig() ReloadResult {
c.reloadLock.Lock() c.reloadLock.Lock()
defer c.reloadLock.Unlock() defer c.reloadLock.Unlock()
c.oldSettings = make(map[string]any) c.oldSettings = make(map[string]any)
maps.Copy(c.oldSettings, c.Settings) maps.Copy(c.oldSettings, c.Settings)
if err := c.Load(c.path); err != nil { err := c.Load(c.path)
c.l.Error("Error occurred while reloading config", "config_path", c.path, "error", err) if err != nil {
return ReloadResult{Err: err, RestartRequired: c.needsRestart} c.l.Error("Error occurred while reloading config",
"config_path", c.path,
"error", err,
)
return
} }
return c.runCallbacks() for _, v := range c.callbacks {
v(c)
}
} }
// ReloadConfigString is ReloadConfig for a config held in memory rather than on disk. func (c *C) ReloadConfigString(raw string) error {
func (c *C) ReloadConfigString(raw string) ReloadResult {
c.reloadLock.Lock() c.reloadLock.Lock()
defer c.reloadLock.Unlock() defer c.reloadLock.Unlock()
c.oldSettings = make(map[string]any) c.oldSettings = make(map[string]any)
maps.Copy(c.oldSettings, c.Settings) maps.Copy(c.oldSettings, c.Settings)
if err := c.LoadString(raw); err != nil { err := c.LoadString(raw)
c.l.Error("Error occurred while reloading config", "error", err) if err != nil {
return ReloadResult{Err: err, RestartRequired: c.needsRestart} return err
} }
return c.runCallbacks() for _, v := range c.callbacks {
} v(c)
func (c *C) runCallbacks() ReloadResult {
r := &Reload{c: c}
for _, v := range c.reloadCallbacks() {
v(c, r)
} }
if r.restart { return nil
c.needsRestart = true
}
return ReloadResult{Failed: r.failed, RestartRequired: c.needsRestart, RestartKeys: r.keys}
} }
// GetString will get the string for k or return the default d if not found or invalid // GetString will get the string for k or return the default d if not found or invalid
@@ -433,7 +391,7 @@ func (c *C) parse() error {
var nm map[string]any var nm map[string]any
err = yaml.Unmarshal(b, &nm) err = yaml.Unmarshal(b, &nm)
if err != nil { if err != nil {
return fmt.Errorf("%s: %w", path, err) return err
} }
// We need to use WithAppendSlice so that firewall rules in separate // We need to use WithAppendSlice so that firewall rules in separate
+2 -41
View File
@@ -3,7 +3,6 @@ package config
import ( import (
"os" "os"
"path/filepath" "path/filepath"
"sync"
"testing" "testing"
"time" "time"
@@ -20,7 +19,7 @@ func TestConfig_Load(t *testing.T) {
// invalid yaml // invalid yaml
c := NewC(l) c := NewC(l)
os.WriteFile(filepath.Join(dir, "01.yaml"), []byte(" invalid yaml"), 0644) os.WriteFile(filepath.Join(dir, "01.yaml"), []byte(" invalid yaml"), 0644)
require.EqualError(t, c.Load(dir), filepath.Join(dir, "01.yaml")+": yaml: unmarshal errors:\n line 1: cannot unmarshal !!str `invalid...` into map[string]interface {}") require.EqualError(t, c.Load(dir), "yaml: unmarshal errors:\n line 1: cannot unmarshal !!str `invalid...` into map[string]interface {}")
// simple multi config merge // simple multi config merge
c = NewC(l) c = NewC(l)
@@ -130,7 +129,7 @@ func TestConfig_ReloadConfig(t *testing.T) {
os.WriteFile(filepath.Join(dir, "01.yaml"), []byte("outer:\n inner: ho"), 0644) os.WriteFile(filepath.Join(dir, "01.yaml"), []byte("outer:\n inner: ho"), 0644)
c.RegisterReloadCallback(func(c *C, _ *Reload) { c.RegisterReloadCallback(func(c *C) {
done <- true done <- true
}) })
@@ -221,41 +220,3 @@ firewall:
} }
assert.Equal(t, expected, m) assert.Equal(t, expected, m)
} }
// TestRegisterReloadCallbackDuringReload: registration is not confined to startup. The
// HUP handler runs from inside nebula.Main, and anything wired after Main returns
// registers while a reload may already be iterating the callbacks. Under -race this is
// the data race on the slice; without it, it checks a callback registered mid-reload is
// picked up by the next one.
func TestRegisterReloadCallbackDuringReload(t *testing.T) {
l := test.NewLogger()
c := NewC(l)
require.NoError(t, c.LoadString("a: 1"))
var wg sync.WaitGroup
wg.Add(2)
go func() {
defer wg.Done()
for i := 0; i < 200; i++ {
c.RegisterReloadCallback(func(*C, *Reload) {})
}
}()
go func() {
defer wg.Done()
for i := 0; i < 200; i++ {
assert.NoError(t, c.ReloadConfigString("a: 1").Err)
}
}()
wg.Wait()
// A callback that registers another callback must not deadlock, and the new one
// runs from the following reload on.
var ran int
c.RegisterReloadCallback(func(c *C, _ *Reload) {
c.RegisterReloadCallback(func(*C, *Reload) { ran++ })
})
require.NoError(t, c.ReloadConfigString("a: 1").Err)
assert.Equal(t, 0, ran, "registered during this reload, so not run by it")
require.NoError(t, c.ReloadConfigString("a: 1").Err)
assert.Equal(t, 1, ran, "and run by the next")
}
-98
View File
@@ -1,98 +0,0 @@
package config
import (
"context"
"fmt"
"log/slog"
"slices"
)
// ReloadResult describes how a reload went. Problems are logged where they happen, Log writes a one line summary.
type ReloadResult struct {
// Err is set when the new config could not be read or parsed, no callbacks ran
Err error
// Failed is true when a callback could not apply part of the new config
Failed bool
// RestartRequired is true once any reload has needed a restart, and stays true until the process restarts
RestartRequired bool
// RestartKeys are the keys this reload changed that only apply on a restart
RestartKeys []string
}
// Summary returns a one line summary of the reload and the level to log it at.
func (r ReloadResult) Summary() (slog.Level, string) {
switch {
case r.Err != nil && r.RestartRequired:
return slog.LevelError, "Config reload failed, the new config did not load and a restart is still required"
case r.Err != nil:
return slog.LevelError, "Config reload failed, the new config did not load"
case r.Failed && r.RestartRequired:
return slog.LevelError, "Config reload completed with errors and requires a restart to fully apply, review the logs"
case r.Failed:
return slog.LevelError, "Config reload completed with errors, review the logs"
case r.RestartRequired:
return slog.LevelWarn, "Config reload was completed but requires a restart to fully apply"
default:
return slog.LevelInfo, "Config reload complete"
}
}
// String is the summary with Err and RestartKeys included, for output other than the log.
func (r ReloadResult) String() string {
_, msg := r.Summary()
if r.Err != nil {
msg = fmt.Sprintf("%s: %v", msg, r.Err)
}
if len(r.RestartKeys) > 0 {
msg = fmt.Sprintf("%s, keys %v", msg, r.RestartKeys)
}
return msg
}
// Log writes Summary to l.
func (r ReloadResult) Log(l *slog.Logger) {
level, msg := r.Summary()
if len(r.RestartKeys) > 0 {
l.Log(context.Background(), level, msg, "keys", r.RestartKeys)
return
}
l.Log(context.Background(), level, msg)
}
// Reload is passed to each reload callback to report how applying the new config went.
type Reload struct {
c *C
failed bool
restart bool
// Changed keys that only apply on a restart, from every callback
keys []string
}
// Failed marks the reload as failed, the caller logs why. Safe to call on a nil Reload, which is what the initial load
// passes.
func (r *Reload) Failed() {
if r != nil {
r.failed = true
}
}
// RestartRequired marks the reload as needing a restart, the caller logs why. Safe to call on a nil Reload.
func (r *Reload) RestartRequired() {
if r != nil {
r.restart = true
}
}
// RestartRequiredIfChanged marks the reload as needing a restart if any of keys changed, and adds them to RestartKeys.
// Use it for keys that are only read at startup. Safe to call on a nil Reload.
func (r *Reload) RestartRequiredIfChanged(keys ...string) {
if r == nil {
return
}
for _, k := range keys {
if r.c.HasChanged(k) && !slices.Contains(r.keys, k) {
r.keys = append(r.keys, k)
r.restart = true
}
}
}
-108
View File
@@ -1,108 +0,0 @@
package config
import (
"errors"
"log/slog"
"testing"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func TestReloadConfigString(t *testing.T) {
c := NewC(test.NewLogger())
require.NoError(t, c.LoadString("a: 1"))
fail := false
ran := 0
c.RegisterReloadCallback(func(c *C, r *Reload) {
ran++
if fail {
r.Failed()
}
})
assert.Equal(t, ReloadResult{}, c.ReloadConfigString("a: 2"))
fail = true
assert.Equal(t, ReloadResult{Failed: true}, c.ReloadConfigString("a: 3"))
assert.Equal(t, 2, ran)
// A config that fails to load runs no callbacks and returns the error
res := c.ReloadConfigString("a: [")
require.Error(t, res.Err)
assert.False(t, res.Failed)
assert.Equal(t, 2, ran)
assert.Equal(t, 3, c.GetInt("a", 0))
}
func TestReloadRestartRequiredLatches(t *testing.T) {
c := NewC(test.NewLogger())
require.NoError(t, c.LoadString("a: 1\nb: 1"))
c.RegisterReloadCallback(func(c *C, r *Reload) { r.RestartRequiredIfChanged("a", "b") })
// Nothing that needs a restart changed
assert.Equal(t, ReloadResult{}, c.ReloadConfigString("a: 1\nb: 1\nc: 1"))
assert.Equal(t, ReloadResult{RestartRequired: true, RestartKeys: []string{"a"}}, c.ReloadConfigString("a: 2\nb: 1\nc: 1"))
// The restart sticks through a reload with no changes, reverting the key and a config that fails to load. Only the
// reload that changed the key lists it
assert.Equal(t, ReloadResult{RestartRequired: true}, c.ReloadConfigString("a: 2\nb: 1\nc: 1"))
assert.Equal(t, ReloadResult{RestartRequired: true, RestartKeys: []string{"a"}}, c.ReloadConfigString("a: 1\nb: 1\nc: 1"))
res := c.ReloadConfigString("a: [")
require.Error(t, res.Err)
assert.True(t, res.RestartRequired)
}
// A key reported by more than one callback is listed once
func TestReloadRestartKeysDeduped(t *testing.T) {
c := NewC(test.NewLogger())
require.NoError(t, c.LoadString("a: 1\nb: 1"))
c.RegisterReloadCallback(func(c *C, r *Reload) { r.RestartRequiredIfChanged("a") })
c.RegisterReloadCallback(func(c *C, r *Reload) { r.RestartRequiredIfChanged("a", "b") })
assert.Equal(t, []string{"a", "b"}, c.ReloadConfigString("a: 2\nb: 2").RestartKeys)
}
// A callback can require a restart without a key changing, a file the config points at may have
func TestReloadRestartRequiredWithoutAChange(t *testing.T) {
c := NewC(test.NewLogger())
require.NoError(t, c.LoadString("a: 1"))
c.RegisterReloadCallback(func(c *C, r *Reload) { r.RestartRequired() })
assert.Equal(t, ReloadResult{RestartRequired: true}, c.ReloadConfigString("a: 1"))
}
func TestReloadResultSummary(t *testing.T) {
cases := []struct {
res ReloadResult
level slog.Level
msg string
}{
{ReloadResult{}, slog.LevelInfo, "Config reload complete"},
{ReloadResult{RestartRequired: true}, slog.LevelWarn, "Config reload was completed but requires a restart to fully apply"},
{ReloadResult{Failed: true}, slog.LevelError, "Config reload completed with errors, review the logs"},
{ReloadResult{Failed: true, RestartRequired: true}, slog.LevelError, "Config reload completed with errors and requires a restart to fully apply, review the logs"},
{ReloadResult{Err: errors.New("bad yaml")}, slog.LevelError, "Config reload failed, the new config did not load"},
{ReloadResult{Err: errors.New("bad yaml"), RestartRequired: true}, slog.LevelError, "Config reload failed, the new config did not load and a restart is still required"},
}
for _, tc := range cases {
level, msg := tc.res.Summary()
assert.Equal(t, tc.level, level, tc.msg)
assert.Equal(t, tc.msg, msg)
}
res := ReloadResult{RestartRequired: true, RestartKeys: []string{"a", "b"}}
assert.Equal(t, "Config reload was completed but requires a restart to fully apply, keys [a b]", res.String())
// String includes the error
assert.Equal(t, "Config reload failed, the new config did not load: bad yaml", ReloadResult{Err: errors.New("bad yaml")}.String())
}
// The initial load passes a nil Reload
func TestReloadNilIsSafe(t *testing.T) {
var r *Reload
r.Failed()
r.RestartRequired()
r.RestartRequiredIfChanged("a")
}
+5 -4
View File
@@ -57,8 +57,7 @@ func newConnectionManagerFromConfig(l *slog.Logger, c *config.C, hm *HostMap, p
} }
cm.reload(c, true) cm.reload(c, true)
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) { c.RegisterReloadCallback(func(c *config.C) {
r.RestartRequiredIfChanged("timers.connection_alive_interval", "timers.pending_deletion_interval")
cm.reload(c, false) cm.reload(c, false)
}) })
@@ -175,8 +174,10 @@ func (cm *connectionManager) doTrafficCheck(localIndex uint32, p, nb, out []byte
switch decision { switch decision {
case deleteTunnel: case deleteTunnel:
// Local delete only, clearing the lighthouse cache if this was the last hostinfo for the vpn addr if cm.hostMap.DeleteHostInfo(hostinfo) {
cm.intf.closeTunnel(hostinfo) // Only clearing the lighthouse cache if this is the last hostinfo for this vpn ip in the hostmap
cm.intf.lightHouse.DeleteVpnAddrs(hostinfo.vpnAddrs)
}
case closeTunnel: case closeTunnel:
cm.intf.sendCloseTunnel(hostinfo) cm.intf.sendCloseTunnel(hostinfo)
-3
View File
@@ -7,7 +7,6 @@ import (
"testing" "testing"
"time" "time"
"github.com/rcrowley/go-metrics"
"github.com/slackhq/nebula/cert" "github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/overlay/overlaytest" "github.com/slackhq/nebula/overlay/overlaytest"
@@ -22,8 +21,6 @@ func newTestLighthouse() *LightHouse {
l: test.NewLogger(), l: test.NewLogger(),
addrMap: map[netip.Addr]*RemoteList{}, addrMap: map[netip.Addr]*RemoteList{},
queryChan: make(chan netip.Addr, 10), queryChan: make(chan netip.Addr, 10),
// Nothing drains queryChan here, so an 11th query is dropped.
queryDropped: metrics.NilCounter{},
} }
lighthouses := []netip.Addr{} lighthouses := []netip.Addr{}
staticList := map[netip.Addr]struct{}{} staticList := map[netip.Addr]struct{}{}
-119
View File
@@ -2,8 +2,6 @@ package nebula
import ( import (
"net/netip" "net/netip"
"strconv"
"sync"
"testing" "testing"
"time" "time"
@@ -13,7 +11,6 @@ import (
ct "github.com/slackhq/nebula/cert_test" ct "github.com/slackhq/nebula/cert_test"
"github.com/slackhq/nebula/handshake" "github.com/slackhq/nebula/handshake"
"github.com/slackhq/nebula/header" "github.com/slackhq/nebula/header"
"github.com/slackhq/nebula/overlay"
"github.com/slackhq/nebula/test" "github.com/slackhq/nebula/test"
"github.com/slackhq/nebula/udp" "github.com/slackhq/nebula/udp"
"github.com/stretchr/testify/assert" "github.com/stretchr/testify/assert"
@@ -150,122 +147,6 @@ func TestSendNoMetricsCloseTunnelKeepsRebindEpoch(t *testing.T) {
assert.True(t, hostinfo.markOut(1), "a CloseTunnel send must not consume the rebind epoch") assert.True(t, hostinfo.markOut(1), "a CloseTunnel send must not consume the rebind epoch")
} }
// capturingConn is a udp.Conn that records every WriteTo.
type capturingConn struct {
udp.NoopConn
writes [][]byte
addrs []netip.AddrPort
}
func (c *capturingConn) WriteTo(b []byte, addr netip.AddrPort) error {
c.writes = append(c.writes, append([]byte(nil), b...))
c.addrs = append(c.addrs, addr)
return nil
}
// TestSendNoMetricsViaRelay sends through a relay from the mtu sized buffer the cached packet flush uses. Payloads up
// to overlay.MaxMTU go out intact, anything bigger is dropped rather than outgrowing the buffer, which once panicked.
func TestSendNoMetricsViaRelay(t *testing.T) {
tests := []struct {
n int
sent bool
}{
{0, true},
{1300, true},
{overlay.MaxMTU, true},
{overlay.MaxMTU + 1, false}, // too big for the relay's tag
{mtu - 47, false}, // too big for the inner tag, this panicked
{9000, false}, // a full packet at tun.mtu 9000
}
for _, tt := range tests {
t.Run(strconv.Itoa(tt.n), func(t *testing.T) {
// One tunnel to the target, carried end to end, and one to the relay that wraps it
targetInit, targetResp := runTestHandshake(t)
relayInit, relayResp := runTestHandshake(t)
ci, err := newConnectionStateFromResult(targetInit)
require.NoError(t, err)
relayCI, err := newConnectionStateFromResult(relayInit)
require.NoError(t, err)
targetAddr := netip.MustParseAddr("10.0.0.2")
relayAddr := netip.MustParseAddr("10.0.0.3")
relayRemote := netip.MustParseAddrPort("192.0.2.1:4242")
relay := &Relay{Type: TerminalType, State: Established, LocalIndex: 5, RemoteIndex: 6, PeerAddr: targetAddr}
relayHI := &HostInfo{
vpnAddrs: []netip.Addr{relayAddr},
ConnectionState: relayCI,
relayState: RelayState{
relayForByAddr: map[netip.Addr]*Relay{targetAddr: relay},
relayForByIdx: map[uint32]*Relay{relay.LocalIndex: relay},
},
}
relayHI.remote.Store(&relayRemote)
// No remote of its own, so the send has to go via the relay
hostinfo := &HostInfo{
vpnAddrs: []netip.Addr{targetAddr},
ConnectionState: ci,
remoteIndexId: 7,
relayState: RelayState{
relays: []netip.Addr{relayAddr},
relayForByAddr: map[netip.Addr]*Relay{},
relayForByIdx: map[uint32]*Relay{},
},
}
l := test.NewLogger()
hm := newHostMap(l)
hm.Hosts[relayAddr] = relayHI
conn := &capturingConn{}
f := &Interface{
l: l,
hostMap: hm,
messageMetrics: &MessageMetrics{txExhausted: metrics.NewCounter()},
writers: []udp.Conn{conn},
}
f.connectionManager = &connectionManager{intf: f, relayUsed: map[uint32]struct{}{}, relayUsedLock: &sync.RWMutex{}}
payload := make([]byte, tt.n)
for i := range payload {
payload[i] = byte(i)
}
counter := ci.messageCounter.Load()
f.sendNoMetrics(header.Message, 0, ci, hostinfo, netip.AddrPort{}, payload, make([]byte, 12), make([]byte, mtu), 0)
if !tt.sent {
assert.Empty(t, conn.writes)
assert.Equal(t, counter, ci.messageCounter.Load(), "a dropped packet must not spend a counter")
return
}
require.Len(t, conn.writes, 1)
assert.Equal(t, relayRemote, conn.addrs[0])
pkt := conn.writes[0]
// The relay authenticates the whole packet under its own header
outer := &header.H{}
require.NoError(t, outer.Parse(pkt))
assert.Equal(t, header.MessageRelay, outer.Subtype)
assert.Equal(t, relay.RemoteIndex, outer.RemoteIndex)
relayPeer, err := newConnectionStateFromResult(relayResp)
require.NoError(t, err)
require.NoError(t, relayPeer.VerifyRelay(l, outer.MessageCounter, pkt, make([]byte, 12)))
// Inside it, the target decrypts the end to end packet back to the payload
inner := pkt[header.Len : len(pkt)-relayPeer.dKey.Overhead()]
h := &header.H{}
require.NoError(t, h.Parse(inner))
assert.Equal(t, header.Message, h.Type)
assert.Equal(t, hostinfo.remoteIndexId, h.RemoteIndex)
targetPeer, err := newConnectionStateFromResult(targetResp)
require.NoError(t, err)
got, err := targetPeer.Decrypt(l, h.MessageCounter, inner, make([]byte, 12))
require.NoError(t, err)
assert.Equal(t, payload, got)
})
}
}
func TestNewConnectionStateFromResult(t *testing.T) { func TestNewConnectionStateFromResult(t *testing.T) {
initR, respR := runTestHandshake(t) initR, respR := runTestHandshake(t)
-9
View File
@@ -11,7 +11,6 @@ import (
"syscall" "syscall"
"github.com/slackhq/nebula/cert" "github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/header" "github.com/slackhq/nebula/header"
"github.com/slackhq/nebula/overlay" "github.com/slackhq/nebula/overlay"
) )
@@ -48,7 +47,6 @@ type Control struct {
f *Interface f *Interface
l *slog.Logger l *slog.Logger
commands *commands.Registry
ctx context.Context ctx context.Context
cancel context.CancelFunc cancel context.CancelFunc
sshStart func() sshStart func()
@@ -374,13 +372,6 @@ func (c *Control) CloseAllTunnels(excludeLighthouses bool) (closed int) {
return return
} }
// Commands returns the registry the daemon's surfaces serve from. It is the live
// registry, not a copy: a command registered on it is served to every ssh session that
// starts afterwards. Clone it to run commands without that reach.
func (c *Control) Commands() *commands.Registry {
return c.commands
}
func (c *Control) Device() overlay.Device { func (c *Control) Device() overlay.Device {
return c.f.inside return c.f.inside
} }
+3 -3
View File
@@ -151,9 +151,9 @@ func (c *fakeConn) WriteTo(_ []byte, _ netip.AddrPort) error { return nil }
func (c *fakeConn) WriteBatch(bufs [][]byte, _ []netip.AddrPort) (int, error) { func (c *fakeConn) WriteBatch(bufs [][]byte, _ []netip.AddrPort) (int, error) {
return len(bufs), nil return len(bufs), nil
} }
func (c *fakeConn) ReloadConfig(_ *config.C, _ *config.Reload) {} func (c *fakeConn) ReloadConfig(_ *config.C) {}
func (c *fakeConn) SupportsMultipleReaders() bool { return true } func (c *fakeConn) SupportsMultipleReaders() bool { return true }
func (c *fakeConn) Close() error { c.closed = true; return nil } func (c *fakeConn) Close() error { c.closed = true; return nil }
type multiqueueDevice struct { type multiqueueDevice struct {
*fakeDevice *fakeDevice
+1 -2
View File
@@ -69,10 +69,9 @@ func newDnsServerFromConfig(ctx context.Context, l *slog.Logger, pki *PKI, hostM
ds.mux = dns.NewServeMux() ds.mux = dns.NewServeMux()
ds.mux.HandleFunc(".", ds.handleDnsRequest) ds.mux.HandleFunc(".", ds.handleDnsRequest)
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) { c.RegisterReloadCallback(func(c *config.C) {
if err := ds.reload(c, false); err != nil { if err := ds.reload(c, false); err != nil {
ds.l.Error("Failed to reload DNS responder from config", "error", err) ds.l.Error("Failed to reload DNS responder from config", "error", err)
r.Failed()
} }
}) })
-169
View File
@@ -1,169 +0,0 @@
//go:build e2e_testing
package e2e
import (
"bytes"
"fmt"
"path/filepath"
"regexp"
"strings"
"testing"
"time"
"github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/cert_test"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/sshd"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/require"
)
// volatile lists everything in command output that legitimately differs between runs.
// Each run mints fresh certificates and indexes, so these are replaced with a marker
// before comparison. Keep the list short: anything added here is something the golden
// stops guarding.
var volatile = []struct {
re *regexp.Regexp
with string
}{
{regexp.MustCompile(`[0-9a-f]{40,}`), "<hex>"}, // fingerprints, keys, signatures
{regexp.MustCompile(`\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}(?:\.\d+)?(?:Z|[+-]\d{2}:\d{2})`), "<time>"}, // RFC 3339
{regexp.MustCompile(`\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}(?:\.\d+)? [+-]\d{4} \w+`), "<time>"}, // time.Time.String()
{regexp.MustCompile(`(?s)(-----BEGIN [A-Z ]+-----\n).*?(\n-----END [A-Z ]+-----)`), "${1}<pem>${2}"}, // print-cert -raw
{regexp.MustCompile(`(-----BEGIN [A-Z ]+-----\\n).*?(\\n-----END [A-Z ]+-----)`), "${1}<pem>${2}"}, // the same, inside a JSON string
{regexp.MustCompile(`"(localIndex|remoteIndex|messageCounter)":\d+`), `"$1":<n>`},
{regexp.MustCompile(`(?m)^(\s*"?(?:LocalIndex|RemoteIndex|localIndex|remoteIndex|messageCounter)"?:?\s*)\d+`), "${1}<n>"}, // and the -pretty form
}
func normalise(s string) string {
for _, v := range volatile {
s = v.re.ReplaceAllString(s, v.with)
}
return s
}
// outputCase is one command line to run through the ssh dispatch path.
type outputCase struct {
name string
args []string
}
// TestCommandOutput pins the bytes every command produces, in every format, through the
// ssh dispatch path. It is the safety net for moving rendering out of the commands: a
// change here is a change someone at an ssh prompt would see. Regenerate with:
//
// go test -tags e2e_testing ./e2e -run TestCommandOutput -update
func TestCommandOutput(t *testing.T) {
ca, _, caKey, _ := cert_test.NewTestCaCert(
cert.Version1, cert.Curve_CURVE25519,
time.Now(), time.Now().Add(10*time.Minute),
nil, nil, []string{},
)
control, _, _, _ := newSimpleServer(cert.Version1, ca, caKey, "output-golden", "10.222.0.1/24", nil)
control.Start()
t.Cleanup(func() { control.Stop() })
// As an ssh session does: a clone of the daemon's registry with logout added, which
// help on the clone lists and describes. There is no session to end here.
reg := control.Commands().Clone()
reg.Register(sshd.LogoutCommand(func() {}))
// Dispatch with ssh's default-format policy, so the no-flag column records what an
// operator at an ssh prompt sees.
reg.DefaultFormat = sshd.DefaultFormat
// run dispatches as the ssh exec path does and returns the output and the error, which
// over ssh is exit status 1.
run := func(args ...string) (string, error) {
buf := &bytes.Buffer{}
err := reg.DispatchArgs(args, commands.MergedOutput(buf))
return buf.String(), err
}
// Seed a pending tunnel so the listing and per-address commands have something to show.
seed, err := run("create-tunnel", "10.222.0.9")
require.NoError(t, err)
require.Equal(t, "Created\n", seed, "seeding a tunnel should succeed before anything else is recorded")
cases := []outputCase{
{"version", nil},
{"device-info", nil},
{"list-hostmap", nil},
{"list-hostmap", []string{"-by-index"}},
{"list-pending-hostmap", nil},
{"list-pending-hostmap", []string{"-by-index"}},
{"list-lighthouse-addrmap", nil},
{"print-cert", nil},
{"print-cert", []string{"-raw"}},
{"print-cert", []string{"10.222.0.9"}},
// A flag after the operand is still a flag, on both surfaces.
{"print-cert", []string{"10.222.0.9", "-json"}},
{"print-cert", []string{"10.222.0.250"}},
{"print-cert", []string{"not-an-address"}},
{"print-tunnel", nil},
{"print-tunnel", []string{"10.222.0.9"}},
{"print-tunnel", []string{"10.222.0.250"}},
{"print-relays", nil},
{"query-lighthouse", nil},
{"query-lighthouse", []string{"10.222.0.9"}},
{"query-lighthouse", []string{"not-an-address"}},
// The test logger cannot change level or format, so these cases record the
// refusal each command prints on it, not a successful change.
{"log-level", nil},
{"log-level", []string{"debug"}},
{"log-level", []string{"bogus"}},
{"log-level", []string{"info"}},
{"log-format", nil},
{"log-format", []string{"json"}},
{"log-format", []string{"text"}},
{"change-remote", nil},
{"change-remote", []string{"10.222.0.9"}},
{"change-remote", []string{"-address", "10.94.0.10:4242", "10.222.0.9"}},
{"change-remote", []string{"-address", "10.94.0.10:4242", "10.222.0.250"}},
{"create-tunnel", nil},
{"create-tunnel", []string{"10.222.0.9"}},
// A fresh address, so this exercises the handshake start with a remote attached
// rather than the "already handshaking" path above.
{"create-tunnel", []string{"-address", "10.94.0.10:4242", "10.222.0.10"}},
{"create-tunnel", []string{"not-an-address"}},
{"close-tunnel", nil},
{"close-tunnel", []string{"10.222.0.250"}},
{"close-tunnel", []string{"not-an-address"}},
{"help", nil},
{"help", []string{"print-cert"}},
{"help", []string{"logout"}},
{"logout", []string{"-h"}},
{"not-a-command", []string{"10.222.0.9"}},
// reload is last so that whatever a reload does to the daemon cannot leak into an
// earlier case. In this harness it fails: the config was loaded from a string and
// has no path to re-read, which is what the recorded result says.
{"reload", nil},
}
got := &strings.Builder{}
for _, c := range cases {
for _, fmtFlags := range [][]string{nil, {"-json"}, {"-pretty"}} {
args := append([]string{c.name}, fmtFlags...)
args = append(args, c.args...)
fmt.Fprintf(got, "$ %s\n", strings.Join(args, " "))
out, err := run(args...)
out = normalise(out)
got.WriteString(out)
// Output that does not end in a newline is recorded as such, the way diff marks
// it, so the golden sees that byte too and the next prompt still starts a line.
if out != "" && !strings.HasSuffix(out, "\n") {
got.WriteString("\\ No newline at end of output\n")
}
// A failed command is recorded as the exit status an ssh caller sees, so the
// golden pins which outcomes count as failures.
if err != nil {
got.WriteString("exit status 1\n")
}
got.WriteString("\n")
}
}
test.Golden(t, filepath.Join("testdata", "commands_output.golden"), got.String())
}
+1 -1
View File
@@ -1513,7 +1513,7 @@ func TestLighthouseUpdateOnReload(t *testing.T) {
// Reload the config. The lighthouse.hosts change triggers TriggerUpdate, // Reload the config. The lighthouse.hosts change triggers TriggerUpdate,
// which wakes the update worker. It calls SendUpdate, initiating a // which wakes the update worker. It calls SendUpdate, initiating a
// handshake to the new lighthouse and caching the HostUpdateNotification. // handshake to the new lighthouse and caching the HostUpdateNotification.
require.NoError(t, myConfig.ReloadConfigString(string(newCfg)).Err) require.NoError(t, myConfig.ReloadConfigString(string(newCfg)))
// Route until the lighthouse receives the HostUpdateNotification. // Route until the lighthouse receives the HostUpdateNotification.
// This covers: handshake stage 1, stage 2, then the cached update. // This covers: handshake stage 1, stage 2, then the cached update.
File diff suppressed because it is too large. Load diff
+2 -2
View File
@@ -129,7 +129,7 @@ func TestCertUpgrade(t *testing.T) {
} }
r.Logf("reload new v2-only config") r.Logf("reload new v2-only config")
err = myC.ReloadConfigString(string(cb)).Err err = myC.ReloadConfigString(string(cb))
assert.NoError(t, err) assert.NoError(t, err)
r.Log("yay, spin until their sees it") r.Log("yay, spin until their sees it")
waitStart := time.Now() waitStart := time.Now()
@@ -222,7 +222,7 @@ func TestCertDowngrade(t *testing.T) {
} }
r.Logf("reload new v1-only config") r.Logf("reload new v1-only config")
err = myC.ReloadConfigString(string(cb)).Err err = myC.ReloadConfigString(string(cb))
assert.NoError(t, err) assert.NoError(t, err)
r.Log("yay, spin until their sees it") r.Log("yay, spin until their sees it")
waitStart := time.Now() waitStart := time.Now()
+25 -50
View File
@@ -8,7 +8,6 @@ pki:
cert: /etc/nebula/host.crt cert: /etc/nebula/host.crt
key: /etc/nebula/host.key key: /etc/nebula/host.key
# blocklist is a list of certificate fingerprints that we will refuse to talk to # blocklist is a list of certificate fingerprints that we will refuse to talk to
# Entries are the hex fingerprints shown by `nebula-cert print`, an invalid entry fails startup and reload.
#blocklist: #blocklist:
# - c99d4e650533b92061b09918e838a5a0a6aaee21eed1d12fd937682865936c72 # - c99d4e650533b92061b09918e838a5a0a6aaee21eed1d12fd937682865936c72
# disconnect_invalid is a toggle to force a client to be disconnected if the certificate is expired or invalid. # disconnect_invalid is a toggle to force a client to be disconnected if the certificate is expired or invalid.
@@ -44,7 +43,7 @@ static_host_map:
lighthouse: lighthouse:
# am_lighthouse is used to enable lighthouse functionality for a node. This should ONLY be true on nodes # am_lighthouse is used to enable lighthouse functionality for a node. This should ONLY be true on nodes
# you have configured to be lighthouses in your network. Not reloadable. # you have configured to be lighthouses in your network
am_lighthouse: false am_lighthouse: false
# serve_dns optionally starts a dns listener that responds to various queries and can even be # serve_dns optionally starts a dns listener that responds to various queries and can even be
# delegated to for resolution # delegated to for resolution
@@ -127,9 +126,8 @@ lighthouse:
# Port Nebula will be listening on. The default here is 4242. For a lighthouse node, the port should be defined, # Port Nebula will be listening on. The default here is 4242. For a lighthouse node, the port should be defined,
# however using port 0 will dynamically assign a port and is recommended for roaming nodes. # however using port 0 will dynamically assign a port and is recommended for roaming nodes.
listen: listen:
# To listen on only ipv4, use "0.0.0.0". Not reloadable. # To listen on only ipv4, use "0.0.0.0"
host: "::" host: "::"
# Not reloadable.
port: 4242 port: 4242
# Sets the max number of packets to pull from the kernel for each syscall (under systems that support recvmmsg) # Sets the max number of packets to pull from the kernel for each syscall (under systems that support recvmmsg)
# default is 64, does not support reload # default is 64, does not support reload
@@ -137,13 +135,12 @@ listen:
# superpacket, so the receive scratch is batch * 64KiB per listening socket (~4MiB per routine at the # superpacket, so the receive scratch is batch * 64KiB per listening socket (~4MiB per routine at the
# default of 64). Lower this to trade peak per-syscall throughput for memory on constrained hosts. # default of 64). Lower this to trade peak per-syscall throughput for memory on constrained hosts.
#batch: 64 #batch: 64
# Linux only. Reloadable, but removing a value leaves the socket at the old size until Nebula restarts.
# Configure socket buffers for the udp side (outside), leave unset to use the system defaults. Values will be doubled by the kernel # Configure socket buffers for the udp side (outside), leave unset to use the system defaults. Values will be doubled by the kernel
# Default is net.core.rmem_default and net.core.wmem_default (/proc/sys/net/core/rmem_default and /proc/sys/net/core/rmem_default) # Default is net.core.rmem_default and net.core.wmem_default (/proc/sys/net/core/rmem_default and /proc/sys/net/core/rmem_default)
# Maximum is limited by memory in the system, SO_RCVBUFFORCE and SO_SNDBUFFORCE is used to avoid having to raise the system wide # Maximum is limited by memory in the system, SO_RCVBUFFORCE and SO_SNDBUFFORCE is used to avoid having to raise the system wide
# max, net.core.rmem_max and net.core.wmem_max # max, net.core.rmem_max and net.core.wmem_max
read_buffer: 3412512 #read_buffer: 10485760
write_buffer: 3412512 #write_buffer: 10485760
# On Windows only # On Windows only
# When true, Nebula installs a WFP (Windows Filtering Platform) PERMIT filter scoped to UDP at the listener port. # When true, Nebula installs a WFP (Windows Filtering Platform) PERMIT filter scoped to UDP at the listener port.
@@ -182,7 +179,7 @@ listen:
# Currently, this defaults to 1 which means we have 1 tun queue reader and 1 # Currently, this defaults to 1 which means we have 1 tun queue reader and 1
# UDP queue reader. Setting this above one will set IFF_MULTI_QUEUE on the tun # UDP queue reader. Setting this above one will set IFF_MULTI_QUEUE on the tun
# device and SO_REUSEPORT on the UDP socket to allow multiple queues. # device and SO_REUSEPORT on the UDP socket to allow multiple queues.
# This option is only supported on Linux. Not reloadable. # This option is only supported on Linux.
#routines: 1 #routines: 1
punchy: punchy:
@@ -206,7 +203,6 @@ punchy:
# Cipher allows you to choose between the available ciphers for your network. Options are chachapoly or aes # Cipher allows you to choose between the available ciphers for your network. Options are chachapoly or aes
# IMPORTANT: this value must be identical on ALL NODES/LIGHTHOUSES. We do not/will not support use of different ciphers simultaneously! # IMPORTANT: this value must be identical on ALL NODES/LIGHTHOUSES. We do not/will not support use of different ciphers simultaneously!
# Not reloadable.
#cipher: aes #cipher: aes
# Preferred ranges is used to define a hint about the local network ranges, which speeds up discovering the fastest # Preferred ranges is used to define a hint about the local network ranges, which speeds up discovering the fastest
@@ -233,26 +229,12 @@ punchy:
# Trusted SSH CA public keys. These are the public keys of the CAs that are allowed to sign SSH keys for access. # Trusted SSH CA public keys. These are the public keys of the CAs that are allowed to sign SSH keys for access.
#trusted_cas: #trusted_cas:
#- "ssh public key string" #- "ssh public key string"
# sandbox_dir restricts file paths for profiling commands (start-cpu-profile, save-heap-profile,
# Local APIs allow accessing running nebula state. # save-mutex-profile) to the specified directory. Relative paths will be resolved within this directory,
#local_api: # and absolute paths outside of it will be rejected. Default is $TMP/nebula-debug.
# There are two types of sockets available: a control socket intended for use on the # The directory is NOT automatically created.
# local machine, primarily for debugging, and an info socket intended to make # Overriding this to "" is the same as "/" and will allow overwriting any path on the host.
# nebula's internal state available to other consumers. #sandbox_dir: /var/tmp/nebula-debug
#
# The control socket is made available by default, and can be accessed via
# `nebula ctl <command>`. It defaults to /var/run/nebula/ctl.sock on Linux, macOS and
# the BSDs; on Windows it is a named pipe under \\.\pipe\ProtectedPrefix\Administrators.
#control:
# Set to false to turn the control socket off. Reloadable.
#enabled: true
# The absolute path where the control socket should live, or a full pipe name on
# Windows. The last directory is created if missing, and must be owned by root or by
# the user running nebula. It must not be writable by anyone else.
#socket: /var/run/nebula/ctl.sock
# A group (or account/group/SID on Windows) that should be allowed to access the
# socket.
#group: nebula
# EXPERIMENTAL: relay support for networks that can't establish direct connections. # EXPERIMENTAL: relay support for networks that can't establish direct connections.
relay: relay:
@@ -271,25 +253,18 @@ relay:
# Configure the private interface. Note: addr is baked into the nebula certificate # Configure the private interface. Note: addr is baked into the nebula certificate
tun: tun:
# When tun is disabled, a lighthouse can be started without a local tun interface (and therefore without root) # When tun is disabled, a lighthouse can be started without a local tun interface (and therefore without root)
# Not reloadable.
disabled: false disabled: false
# Name of the device. If not set, a default will be chosen by the OS. # Name of the device. If not set, a default will be chosen by the OS.
# For macOS: if set, must be in the form `utun[0-9]+`. # For macOS: if set, must be in the form `utun[0-9]+`.
# For OpenBSD and NetBSD: if set, must be in the form `tun[0-9]+`, otherwise the first free one is used. # For NetBSD: Required to be set, must be in the form `tun[0-9]+`
# Not reloadable.
dev: nebula1 dev: nebula1
# Toggles forwarding of local broadcast packets, the address of which depends on the ip/mask encoded in pki.cert # Toggles forwarding of local broadcast packets, the address of which depends on the ip/mask encoded in pki.cert
# Not reloadable.
drop_local_broadcast: false drop_local_broadcast: false
# Toggles forwarding of multicast packets. Not reloadable. # Toggles forwarding of multicast packets
drop_multicast: false drop_multicast: false
# Sets the transmit queue length, if you notice lots of transmit drops on the tun it may help to raise this number. Default is 5000 # Sets the transmit queue length, if you notice lots of transmit drops on the tun it may help to raise this number. Default is 500
# Not reloadable. tx_queue: 500
tx_queue: 5000
# Default MTU for every packet, safe setting is (and the default) 1300 for internet based traffic # Default MTU for every packet, safe setting is (and the default) 1300 for internet based traffic
# Anything above 8937, here or on a route below, is capped to 8937 with a warning
# Reloadable on Linux only. Where the tun is handed to nebula by a host app (iOS, Android, macOS network extensions)
# the host app owns the MTU.
mtu: 1300 mtu: 1300
# the use_offloads setting controls if Nebula will attempt to enable GSO and GRO for the tun device. Linux only, not reloadable. # the use_offloads setting controls if Nebula will attempt to enable GSO and GRO for the tun device. Linux only, not reloadable.
@@ -351,9 +326,6 @@ tun:
# `mtu`: will default to tun mtu if this option is not specified # `mtu`: will default to tun mtu if this option is not specified
# `metric`: will default to 0 if this option is not specified # `metric`: will default to 0 if this option is not specified
# `install`: will default to true, controls whether this route is installed in the systems routing table. # `install`: will default to true, controls whether this route is installed in the systems routing table.
# On macOS, a route whose destination another interface already routes, like the local network, is installed once
# that route is removed. One deleted by hand stays deleted until nebula reloads or restarts, or until another
# interface routes its destination and lets it go.
# This setting is reloadable. # This setting is reloadable.
unsafe_routes: unsafe_routes:
#- route: 172.16.1.0/24 #- route: 172.16.1.0/24
@@ -385,7 +357,7 @@ tun:
#use_system_route_table: false #use_system_route_table: false
# Buffer size for reading routes updates. 0 means default system buffer size. (/proc/sys/net/core/rmem_default). # Buffer size for reading routes updates. 0 means default system buffer size. (/proc/sys/net/core/rmem_default).
# If using massive routes updates, for example BGP, you may need to increase this value to avoid packet loss. # If using massive routes updates, for example BGP, you may need to increase this value to avoid packet loss.
# SO_RCVBUFFORCE is used to avoid having to raise the system wide max. Not reloadable. # SO_RCVBUFFORCE is used to avoid having to raise the system wide max
#use_system_route_table_buffer_size: 0 #use_system_route_table_buffer_size: 0
# Configure logging level # Configure logging level
@@ -435,16 +407,14 @@ logging:
# Handshakes are sent to all known addresses at each interval with a linear backoff, # Handshakes are sent to all known addresses at each interval with a linear backoff,
# Wait try_interval after the 1st attempt, 2 * try_interval after the 2nd, etc, until the handshake is older than timeout # Wait try_interval after the 1st attempt, 2 * try_interval after the 2nd, etc, until the handshake is older than timeout
# A 100ms interval with the default 10 retries will give a handshake 5.5 seconds to resolve before timing out # A 100ms interval with the default 10 retries will give a handshake 5.5 seconds to resolve before timing out
# Not reloadable.
#try_interval: 100ms #try_interval: 100ms
#retries: 20 #retries: 20
# query_buffer is the size of the buffer channel for querying lighthouses, drops are recorded in the lighthouse.query.dropped metric. # query_buffer is the size of the buffer channel for querying lighthouses
# Not reloadable. #query_buffer: 64
#query_buffer: 1024
# trigger_buffer is the size of the buffer channel for quickly sending handshakes # trigger_buffer is the size of the buffer channel for quickly sending handshakes
# after receiving the response for lighthouse queries. Not reloadable. # after receiving the response for lighthouse queries
#trigger_buffer: 64 #trigger_buffer: 64
# Tunnel manager settings # Tunnel manager settings
@@ -479,7 +449,9 @@ firewall:
#default_local_cidr_any: false #default_local_cidr_any: false
conntrack: conntrack:
# tcp_timeout also applies to sctp and dccp
tcp_timeout: 12m tcp_timeout: 12m
# udp_timeout also applies to udplite
udp_timeout: 3m udp_timeout: 3m
default_timeout: 10m default_timeout: 10m
@@ -487,8 +459,11 @@ firewall:
# Rules are comprised of a protocol, port, and one or more of host, group, or CIDR # Rules are comprised of a protocol, port, and one or more of host, group, or CIDR
# Logical evaluation is roughly: port AND proto AND (ca_sha OR ca_name) AND (host OR group OR groups OR cidr) AND (local cidr) # Logical evaluation is roughly: port AND proto AND (ca_sha OR ca_name) AND (host OR group OR groups OR cidr) AND (local cidr)
# - port: Takes `0` or `any` as any, a single number `80`, a range `200-901`, or `fragment` to match second and further fragments of fragmented packets (since there is no port available). # - port: Takes `0` or `any` as any, a single number `80`, a range `200-901`, or `fragment` to match second and further fragments of fragmented packets (since there is no port available).
# proto: `any`, `tcp`, `udp`, or `icmp` # proto: `any`, `tcp`, `udp`, `udplite`, `dccp`, `sctp`, `icmp`, or any other IP protocol number from 1 to 255, ie `47` for gre
# a port specification is ignored if proto is `icmp` # a port specification is ignored if proto is `icmp`
# `icmp`, `1`, and `58` are the same, each matches both ICMP and ICMPv6
# Nebula only reads ports from tcp, udp, udplite, dccp, and sctp, so a proto `any` rule with a specific port or range only matches those protocols.
# Rules for other protocols (gre, esp, etc) need port `any`.
# host: `any` or a literal hostname, ie `test-host` # host: `any` or a literal hostname, ie `test-host`
# group: `any` or a literal group name, ie `default-group` # group: `any` or a literal group name, ie `default-group`
# groups: Same as group but accepts a list of values. Multiple values are AND'd together and a certificate would have to contain all groups to pass # groups: Same as group but accepts a list of values. Multiple values are AND'd together and a certificate would have to contain all groups to pass
+41 -778
View File
@@ -2,17 +2,10 @@ package nebula
import ( import (
"context" "context"
"crypto/sha256"
"encoding/hex"
"errors" "errors"
"fmt"
"hash/fnv"
"log/slog" "log/slog"
"net/netip" "net/netip"
"reflect"
"slices"
"strconv" "strconv"
"strings"
"sync" "sync"
"time" "time"
@@ -24,10 +17,6 @@ import (
"github.com/slackhq/nebula/iputil" "github.com/slackhq/nebula/iputil"
) )
type FirewallInterface interface {
AddRule(incoming bool, proto uint8, startPort int32, endPort int32, groups []string, host string, cidr, localCidr string, caName string, caSha string) error
}
type conn struct { type conn struct {
Expires time.Time // Time when this conntrack entry will expire Expires time.Time // Time when this conntrack entry will expire
@@ -42,8 +31,7 @@ type conn struct {
type Firewall struct { type Firewall struct {
Conntrack *FirewallConntrack Conntrack *FirewallConntrack
InRules *FirewallTable rules firewall.Rules
OutRules *FirewallTable
InboundSendReject bool InboundSendReject bool
OutboundSendReject bool OutboundSendReject bool
@@ -63,15 +51,10 @@ type Firewall struct {
// unsafeNetworks is the list of unsafe networks issued to us in the certificate // unsafeNetworks is the list of unsafe networks issued to us in the certificate
unsafeNetworks []netip.Prefix unsafeNetworks []netip.Prefix
rules string
rulesVersion uint16 rulesVersion uint16
// source is the rendered firewall config this firewall was built from, the cert networks are kept above incomingMetrics firewallMetrics
source string outgoingMetrics firewallMetrics
defaultLocalCIDRAny bool
incomingMetrics firewallMetrics
outgoingMetrics firewallMetrics
l *slog.Logger l *slog.Logger
} }
@@ -89,52 +72,6 @@ type FirewallConntrack struct {
TimerWheel *TimerWheel[firewall.Packet] TimerWheel *TimerWheel[firewall.Packet]
} }
// FirewallTable is the entry point for a rule, the evaluation order is:
// Proto AND port AND (CA SHA or CA name) AND local CIDR AND (group OR groups OR name OR remote CIDR)
type FirewallTable struct {
TCP firewallPort
UDP firewallPort
ICMP firewallPort
AnyProto firewallPort
}
func newFirewallTable() *FirewallTable {
return &FirewallTable{
TCP: firewallPort{},
UDP: firewallPort{},
ICMP: firewallPort{},
AnyProto: firewallPort{},
}
}
type FirewallCA struct {
Any *FirewallRule
CANames map[string]*FirewallRule
CAShas map[string]*FirewallRule
}
type FirewallRule struct {
// Any makes Hosts, Groups, and CIDR irrelevant
Any *firewallLocalCIDR
Hosts map[string]*firewallLocalCIDR
Groups []*firewallGroups
CIDR *bart.Table[*firewallLocalCIDR]
}
type firewallGroups struct {
Groups []string
LocalCIDR *firewallLocalCIDR
}
// Even though ports are uint16, int32 maps are faster for lookup
// Plus we can use `-1` for fragment rules
type firewallPort map[int32]*FirewallCA
type firewallLocalCIDR struct {
Any bool
LocalCIDR *bart.Lite
}
// NewFirewall creates a new Firewall object. A TimerWheel is created for you from the provided timeouts. // NewFirewall creates a new Firewall object. A TimerWheel is created for you from the provided timeouts.
// The certificate provided should be the highest version loaded in memory. // The certificate provided should be the highest version loaded in memory.
func NewFirewall(l *slog.Logger, tcpTimeout, UDPTimeout, defaultTimeout time.Duration, c cert.Certificate) *Firewall { func NewFirewall(l *slog.Logger, tcpTimeout, UDPTimeout, defaultTimeout time.Duration, c cert.Certificate) *Firewall {
@@ -173,8 +110,8 @@ func NewFirewall(l *slog.Logger, tcpTimeout, UDPTimeout, defaultTimeout time.Dur
Conns: make(map[firewall.Packet]*conn), Conns: make(map[firewall.Packet]*conn),
TimerWheel: NewTimerWheel[firewall.Packet](tmin, tmax), TimerWheel: NewTimerWheel[firewall.Packet](tmin, tmax),
}, },
InRules: newFirewallTable(), // Nothing is allowed until there are rules
OutRules: newFirewallTable(), rules: firewall.NewRulesBuilder(l).Build(assignedNetworks, unsafeNetworks),
TCPTimeout: tcpTimeout, TCPTimeout: tcpTimeout,
UDPTimeout: UDPTimeout, UDPTimeout: UDPTimeout,
DefaultTimeout: defaultTimeout, DefaultTimeout: defaultTimeout,
@@ -206,31 +143,15 @@ func NewFirewallFromConfig(l *slog.Logger, cs *CertState, c *config.C) (*Firewal
panic("No certificate available to reconfigure the firewall") panic("No certificate available to reconfigure the firewall")
} }
tcpTimeout := c.GetDuration("firewall.conntrack.tcp_timeout", time.Minute*12)
udpTimeout := c.GetDuration("firewall.conntrack.udp_timeout", time.Minute*3)
defaultTimeout := c.GetDuration("firewall.conntrack.default_timeout", time.Minute*10)
if tcpTimeout <= 0 {
return nil, fmt.Errorf("firewall.conntrack.tcp_timeout must be greater than 0, got: %s", tcpTimeout)
}
if udpTimeout <= 0 {
return nil, fmt.Errorf("firewall.conntrack.udp_timeout must be greater than 0, got: %s", udpTimeout)
}
if defaultTimeout <= 0 {
return nil, fmt.Errorf("firewall.conntrack.default_timeout must be greater than 0, got: %s", defaultTimeout)
}
fw := NewFirewall( fw := NewFirewall(
l, l,
tcpTimeout, c.GetDuration("firewall.conntrack.tcp_timeout", time.Minute*12),
udpTimeout, c.GetDuration("firewall.conntrack.udp_timeout", time.Minute*3),
defaultTimeout, c.GetDuration("firewall.conntrack.default_timeout", time.Minute*10),
certificate, certificate,
//TODO: max_connections //TODO: max_connections
) )
fw.defaultLocalCIDRAny = c.GetBool("firewall.default_local_cidr_any", false)
inboundAction := c.GetString("firewall.inbound_action", "drop") inboundAction := c.GetString("firewall.inbound_action", "drop")
switch inboundAction { switch inboundAction {
case "reject": case "reject":
@@ -253,81 +174,23 @@ func NewFirewallFromConfig(l *slog.Logger, cs *CertState, c *config.C) (*Firewal
fw.OutboundSendReject = false fw.OutboundSendReject = false
} }
err := AddFirewallRulesFromConfig(l, false, c, fw) rules, err := firewall.RulesFromConfig(l, c, fw.assignedNetworks, fw.unsafeNetworks)
if err != nil { if err != nil {
return nil, err return nil, err
} }
fw.rules = rules
err = AddFirewallRulesFromConfig(l, true, c, fw)
if err != nil {
return nil, err
}
fw.source = c.Render("firewall")
return fw, nil return fw, nil
} }
// AddRule properly creates the in memory rule structure for a firewall table.
func (f *Firewall) AddRule(incoming bool, proto uint8, startPort int32, endPort int32, groups []string, host string, cidr, localCidr, caName string, caSha string) error {
var (
ft *FirewallTable
fp firewallPort
)
if incoming {
ft = f.InRules
} else {
ft = f.OutRules
}
switch proto {
case iputil.IPProtocolTCP:
fp = ft.TCP
case iputil.IPProtocolUDP:
fp = ft.UDP
case iputil.IPProtocolICMP, iputil.IPProtocolICMPv6:
//ICMP traffic doesn't have ports, so we always coerce to "any", even if a value is provided
if startPort != firewall.PortAny {
f.l.Warn("ignoring port specification for ICMP firewall rule", "startPort", startPort)
}
startPort = firewall.PortAny
endPort = firewall.PortAny
fp = ft.ICMP
case firewall.ProtoAny:
fp = ft.AnyProto
default:
return fmt.Errorf("unknown protocol %v", proto)
}
// We need this rule string because we generate a hash. Removing this will break firewall reload.
ruleString := fmt.Sprintf(
"incoming: %v, proto: %v, startPort: %v, endPort: %v, groups: %v, host: %v, ip: %v, localIp: %v, caName: %v, caSha: %s",
incoming, proto, startPort, endPort, groups, host, cidr, localCidr, caName, caSha,
)
f.rules += ruleString + "\n"
direction := "incoming"
if !incoming {
direction = "outgoing"
}
f.l.Info("Firewall rule added",
"firewallRule", m{"direction": direction, "proto": proto, "startPort": startPort, "endPort": endPort, "groups": groups, "host": host, "cidr": cidr, "localCidr": localCidr, "caName": caName, "caSha": caSha},
)
return fp.addRule(f, startPort, endPort, groups, host, cidr, localCidr, caName, caSha)
}
// GetRuleHash returns a hash representation of all inbound and outbound rules // GetRuleHash returns a hash representation of all inbound and outbound rules
func (f *Firewall) GetRuleHash() string { func (f *Firewall) GetRuleHash() string {
sum := sha256.Sum256([]byte(f.rules)) return f.rules.Hash
return hex.EncodeToString(sum[:])
} }
// GetRuleHashFNV returns a uint32 FNV-1 hash representation the rules, for use as a metric value // GetRuleHashFNV returns a uint32 FNV-1 hash representation the rules, for use as a metric value
func (f *Firewall) GetRuleHashFNV() uint32 { func (f *Firewall) GetRuleHashFNV() uint32 {
h := fnv.New32a() return f.rules.HashFNV
h.Write([]byte(f.rules))
return h.Sum32()
} }
// GetRuleHashes returns both the sha256 and FNV-1 hashes, suitable for logging // GetRuleHashes returns both the sha256 and FNV-1 hashes, suitable for logging
@@ -335,104 +198,6 @@ func (f *Firewall) GetRuleHashes() string {
return "SHA:" + f.GetRuleHash() + ",FNV:" + strconv.FormatUint(uint64(f.GetRuleHashFNV()), 10) return "SHA:" + f.GetRuleHash() + ",FNV:" + strconv.FormatUint(uint64(f.GetRuleHashFNV()), 10)
} }
func AddFirewallRulesFromConfig(l *slog.Logger, inbound bool, c *config.C, fw FirewallInterface) error {
var table string
if inbound {
table = "firewall.inbound"
} else {
table = "firewall.outbound"
}
r := c.Get(table)
if r == nil {
return nil
}
rs, ok := r.([]any)
if !ok {
return fmt.Errorf("%s failed to parse, should be an array of rules", table)
}
for i, t := range rs {
r, err := convertRule(l, t, table, i)
if err != nil {
return fmt.Errorf("%s rule #%v; %s", table, i, err)
}
if r.Code != "" && r.Port != "" {
return fmt.Errorf("%s rule #%v; only one of port or code should be provided", table, i)
}
if r.Host == "" && len(r.Groups) == 0 && r.Cidr == "" && r.LocalCidr == "" && r.CAName == "" && r.CASha == "" {
return fmt.Errorf("%s rule #%v; at least one of host, group, cidr, local_cidr, ca_name, or ca_sha must be provided", table, i)
}
var sPort, errPort string
if r.Code != "" {
errPort = "code"
sPort = r.Code
} else {
errPort = "port"
sPort = r.Port
}
var proto uint8
var startPort, endPort int32
switch r.Proto {
case "any":
proto = firewall.ProtoAny
startPort, endPort, err = parsePort(sPort)
case "tcp":
proto = iputil.IPProtocolTCP
startPort, endPort, err = parsePort(sPort)
case "udp":
proto = iputil.IPProtocolUDP
startPort, endPort, err = parsePort(sPort)
case "icmp":
proto = iputil.IPProtocolICMP
startPort = firewall.PortAny
endPort = firewall.PortAny
if sPort != "" {
l.Warn("ignoring port specification for ICMP firewall rule", "port", sPort)
}
default:
return fmt.Errorf("%s rule #%v; proto was not understood; `%s`", table, i, r.Proto)
}
if err != nil {
return fmt.Errorf("%s rule #%v; %s %s", table, i, errPort, err)
}
if r.Cidr != "" && r.Cidr != "any" {
_, err = netip.ParsePrefix(r.Cidr)
if err != nil {
return fmt.Errorf("%s rule #%v; cidr did not parse; %s", table, i, err)
}
}
if r.LocalCidr != "" && r.LocalCidr != "any" {
_, err = netip.ParsePrefix(r.LocalCidr)
if err != nil {
return fmt.Errorf("%s rule #%v; local_cidr did not parse; %s", table, i, err)
}
}
if warning := r.sanity(); warning != nil {
l.Warn("firewall rule sanity check",
"table", table,
"rule", i,
"warning", warning,
)
}
err = fw.AddRule(inbound, proto, startPort, endPort, r.Groups, r.Host, r.Cidr, r.LocalCidr, r.CAName, r.CASha)
if err != nil {
return fmt.Errorf("%s rule #%v; `%s`", table, i, err)
}
}
return nil
}
var ErrUnknownNetworkType = errors.New("unknown network type") var ErrUnknownNetworkType = errors.New("unknown network type")
var ErrPeerRejected = errors.New("remote address is not within a network that we handle") var ErrPeerRejected = errors.New("remote address is not within a network that we handle")
var ErrInvalidRemoteIP = errors.New("remote address is not in remote certificate networks") var ErrInvalidRemoteIP = errors.New("remote address is not in remote certificate networks")
@@ -476,23 +241,17 @@ func (f *Firewall) Drop(fp firewall.Packet, incoming bool, h *HostInfo, caPool *
} }
// Check if we spoke to this tuple, if we did then allow this packet // Check if we spoke to this tuple, if we did then allow this packet
if f.inConns(fp, h, caPool, localCache) { if f.inConns(&fp, h, caPool, localCache) {
return nil return nil
} }
table := f.OutRules if !f.rules.Match(&fp, incoming, h.ConnectionState.peerCert, caPool) {
if incoming {
table = f.InRules
}
// We now know which firewall table to check against
if !table.match(fp, incoming, h.ConnectionState.peerCert, caPool) {
f.metrics(incoming).droppedNoRule.Inc(1) f.metrics(incoming).droppedNoRule.Inc(1)
return ErrNoMatchingRule return ErrNoMatchingRule
} }
// We always want to conntrack since it is a faster operation // We always want to conntrack since it is a faster operation
f.addConn(fp, incoming) f.addConn(&fp, incoming)
return nil return nil
} }
@@ -521,9 +280,9 @@ func (f *Firewall) EmitStats() {
metrics.GetOrRegisterGauge("firewall.rules.hash", nil).Update(int64(f.GetRuleHashFNV())) metrics.GetOrRegisterGauge("firewall.rules.hash", nil).Update(int64(f.GetRuleHashFNV()))
} }
func (f *Firewall) inConns(fp firewall.Packet, h *HostInfo, caPool *cert.CAPool, localCache firewall.ConntrackCache) bool { func (f *Firewall) inConns(fp *firewall.Packet, h *HostInfo, caPool *cert.CAPool, localCache firewall.ConntrackCache) bool {
if localCache != nil { if localCache != nil {
if _, ok := localCache[fp]; ok { if _, ok := localCache[*fp]; ok {
return true return true
} }
} }
@@ -536,7 +295,7 @@ func (f *Firewall) inConns(fp firewall.Packet, h *HostInfo, caPool *cert.CAPool,
f.evict(ep) f.evict(ep)
} }
c, ok := conntrack.Conns[fp] c, ok := conntrack.Conns[*fp]
if !ok { if !ok {
conntrack.Unlock() conntrack.Unlock()
@@ -546,29 +305,23 @@ func (f *Firewall) inConns(fp firewall.Packet, h *HostInfo, caPool *cert.CAPool,
if c.rulesVersion != f.rulesVersion { if c.rulesVersion != f.rulesVersion {
// This conntrack entry was for an older rule set, validate // This conntrack entry was for an older rule set, validate
// it still passes with the current rule set // it still passes with the current rule set
table := f.OutRules if !f.rules.Match(fp, c.incoming, h.ConnectionState.peerCert, caPool) {
if c.incoming {
table = f.InRules
}
// We now know which firewall table to check against
if !table.match(fp, c.incoming, h.ConnectionState.peerCert, caPool) {
if f.l.Enabled(context.Background(), slog.LevelDebug) { if f.l.Enabled(context.Background(), slog.LevelDebug) {
h.logger(f.l).Debug("dropping old conntrack entry, does not match new ruleset", h.logger(f.l).Debug("dropping old conntrack entry, does not match new ruleset",
"fwPacket", fp, "fwPacket", *fp,
"incoming", c.incoming, "incoming", c.incoming,
"rulesVersion", f.rulesVersion, "rulesVersion", f.rulesVersion,
"oldRulesVersion", c.rulesVersion, "oldRulesVersion", c.rulesVersion,
) )
} }
delete(conntrack.Conns, fp) delete(conntrack.Conns, *fp)
conntrack.Unlock() conntrack.Unlock()
return false return false
} }
if f.l.Enabled(context.Background(), slog.LevelDebug) { if f.l.Enabled(context.Background(), slog.LevelDebug) {
h.logger(f.l).Debug("keeping old conntrack entry, does match new ruleset", h.logger(f.l).Debug("keeping old conntrack entry, does match new ruleset",
"fwPacket", fp, "fwPacket", *fp,
"incoming", c.incoming, "incoming", c.incoming,
"rulesVersion", f.rulesVersion, "rulesVersion", f.rulesVersion,
"oldRulesVersion", c.rulesVersion, "oldRulesVersion", c.rulesVersion,
@@ -578,42 +331,39 @@ func (f *Firewall) inConns(fp firewall.Packet, h *HostInfo, caPool *cert.CAPool,
c.rulesVersion = f.rulesVersion c.rulesVersion = f.rulesVersion
} }
switch fp.Protocol { c.Expires = time.Now().Add(f.conntrackTimeout(fp.Protocol))
case iputil.IPProtocolTCP:
c.Expires = time.Now().Add(f.TCPTimeout)
case iputil.IPProtocolUDP:
c.Expires = time.Now().Add(f.UDPTimeout)
default:
c.Expires = time.Now().Add(f.DefaultTimeout)
}
conntrack.Unlock() conntrack.Unlock()
if localCache != nil { if localCache != nil {
localCache[fp] = struct{}{} localCache[*fp] = struct{}{}
} }
return true return true
} }
func (f *Firewall) addConn(fp firewall.Packet, incoming bool) { // conntrackTimeout returns how long a conntrack entry for proto lives without traffic.
var timeout time.Duration // Connection oriented protocols share the tcp timeout, datagram protocols share the udp timeout.
c := &conn{} func (f *Firewall) conntrackTimeout(proto uint8) time.Duration {
switch proto {
switch fp.Protocol { case iputil.IPProtocolTCP, iputil.IPProtocolSCTP, iputil.IPProtocolDCCP:
case iputil.IPProtocolTCP: return f.TCPTimeout
timeout = f.TCPTimeout case iputil.IPProtocolUDP, iputil.IPProtocolUDPLite:
case iputil.IPProtocolUDP: return f.UDPTimeout
timeout = f.UDPTimeout
default: default:
timeout = f.DefaultTimeout return f.DefaultTimeout
} }
}
func (f *Firewall) addConn(fp *firewall.Packet, incoming bool) {
timeout := f.conntrackTimeout(fp.Protocol)
c := &conn{}
conntrack := f.Conntrack conntrack := f.Conntrack
conntrack.Lock() conntrack.Lock()
if _, ok := conntrack.Conns[fp]; !ok { if _, ok := conntrack.Conns[*fp]; !ok {
conntrack.TimerWheel.Advance(time.Now()) conntrack.TimerWheel.Advance(time.Now())
conntrack.TimerWheel.Add(fp, timeout) conntrack.TimerWheel.Add(*fp, timeout)
} }
// Record which rulesVersion allowed this connection, so we can retest after // Record which rulesVersion allowed this connection, so we can retest after
@@ -621,7 +371,7 @@ func (f *Firewall) addConn(fp firewall.Packet, incoming bool) {
c.incoming = incoming c.incoming = incoming
c.rulesVersion = f.rulesVersion c.rulesVersion = f.rulesVersion
c.Expires = time.Now().Add(timeout) c.Expires = time.Now().Add(timeout)
conntrack.Conns[fp] = c conntrack.Conns[*fp] = c
conntrack.Unlock() conntrack.Unlock()
} }
@@ -647,490 +397,3 @@ func (f *Firewall) evict(p firewall.Packet) {
// This conn is done // This conn is done
delete(conntrack.Conns, p) delete(conntrack.Conns, p)
} }
func (ft *FirewallTable) match(p firewall.Packet, incoming bool, c *cert.CachedCertificate, caPool *cert.CAPool) bool {
if ft.AnyProto.match(p, incoming, c, caPool) {
return true
}
switch p.Protocol {
case iputil.IPProtocolTCP:
if ft.TCP.match(p, incoming, c, caPool) {
return true
}
case iputil.IPProtocolUDP:
if ft.UDP.match(p, incoming, c, caPool) {
return true
}
case iputil.IPProtocolICMP, iputil.IPProtocolICMPv6:
if ft.ICMP.match(p, incoming, c, caPool) {
return true
}
}
return false
}
func (fp firewallPort) addRule(f *Firewall, startPort int32, endPort int32, groups []string, host string, cidr, localCidr, caName string, caSha string) error {
if startPort > endPort {
return fmt.Errorf("start port was lower than end port")
}
for i := startPort; i <= endPort; i++ {
if _, ok := fp[i]; !ok {
fp[i] = &FirewallCA{
CANames: make(map[string]*FirewallRule),
CAShas: make(map[string]*FirewallRule),
}
}
if err := fp[i].addRule(f, groups, host, cidr, localCidr, caName, caSha); err != nil {
return err
}
}
return nil
}
func (fp firewallPort) match(p firewall.Packet, incoming bool, c *cert.CachedCertificate, caPool *cert.CAPool) bool {
// We don't have any allowed ports, bail
if fp == nil {
return false
}
// this branch is here to catch traffic from FirewallTable.Any.match and FirewallTable.ICMP.match
if p.Protocol == iputil.IPProtocolICMP || p.Protocol == iputil.IPProtocolICMPv6 {
// port numbers are re-used for connection tracking of ICMP,
// but we don't want to actually filter on them.
return fp[firewall.PortAny].match(p, c, caPool)
}
var port int32
if p.Fragment {
port = firewall.PortFragment
} else if incoming {
port = int32(p.LocalPort)
} else {
port = int32(p.RemotePort)
}
if fp[port].match(p, c, caPool) {
return true
}
return fp[firewall.PortAny].match(p, c, caPool)
}
func (fc *FirewallCA) addRule(f *Firewall, groups []string, host string, cidr, localCidr, caName, caSha string) error {
fr := func() *FirewallRule {
return &FirewallRule{
Hosts: make(map[string]*firewallLocalCIDR),
Groups: make([]*firewallGroups, 0),
CIDR: new(bart.Table[*firewallLocalCIDR]),
}
}
if caSha == "" && caName == "" {
if fc.Any == nil {
fc.Any = fr()
}
return fc.Any.addRule(f, groups, host, cidr, localCidr)
}
if caSha != "" {
if _, ok := fc.CAShas[caSha]; !ok {
fc.CAShas[caSha] = fr()
}
err := fc.CAShas[caSha].addRule(f, groups, host, cidr, localCidr)
if err != nil {
return err
}
}
if caName != "" {
if _, ok := fc.CANames[caName]; !ok {
fc.CANames[caName] = fr()
}
err := fc.CANames[caName].addRule(f, groups, host, cidr, localCidr)
if err != nil {
return err
}
}
return nil
}
func (fc *FirewallCA) match(p firewall.Packet, c *cert.CachedCertificate, caPool *cert.CAPool) bool {
if fc == nil {
return false
}
if fc.Any.match(p, c) {
return true
}
if t, ok := fc.CAShas[c.Certificate.Issuer()]; ok {
if t.match(p, c) {
return true
}
}
s, err := caPool.GetCAForCert(c.Certificate)
if err != nil {
return false
}
return fc.CANames[s.Certificate.Name()].match(p, c)
}
func (fr *FirewallRule) addRule(f *Firewall, groups []string, host, cidr, localCidr string) error {
flc := func() *firewallLocalCIDR {
return &firewallLocalCIDR{
LocalCIDR: new(bart.Lite),
}
}
if fr.isAny(groups, host, cidr) {
if fr.Any == nil {
fr.Any = flc()
}
return fr.Any.addRule(f, localCidr)
}
if len(groups) > 0 {
nlc := flc()
err := nlc.addRule(f, localCidr)
if err != nil {
return err
}
fr.Groups = append(fr.Groups, &firewallGroups{
Groups: groups,
LocalCIDR: nlc,
})
}
if host != "" {
nlc := fr.Hosts[host]
if nlc == nil {
nlc = flc()
}
err := nlc.addRule(f, localCidr)
if err != nil {
return err
}
fr.Hosts[host] = nlc
}
if cidr != "" {
c, err := netip.ParsePrefix(cidr)
if err != nil {
return err
}
nlc, _ := fr.CIDR.Get(c)
if nlc == nil {
nlc = flc()
}
err = nlc.addRule(f, localCidr)
if err != nil {
return err
}
fr.CIDR.Insert(c, nlc)
}
return nil
}
func (fr *FirewallRule) isAny(groups []string, host string, cidr string) bool {
if len(groups) == 0 && host == "" && cidr == "" {
return true
}
if slices.Contains(groups, "any") {
return true
}
if host == "any" {
return true
}
if cidr == "any" {
return true
}
return false
}
func (fr *FirewallRule) match(p firewall.Packet, c *cert.CachedCertificate) bool {
if fr == nil {
return false
}
// Shortcut path for if groups, hosts, or cidr contained an `any`
if fr.Any.match(p, c) {
return true
}
// Need any of group, host, or cidr to match
for _, sg := range fr.Groups {
found := false
for _, g := range sg.Groups {
if _, ok := c.InvertedGroups[g]; !ok {
found = false
break
}
found = true
}
if found && sg.LocalCIDR.match(p, c) {
return true
}
}
if fr.Hosts != nil {
if flc, ok := fr.Hosts[c.Certificate.Name()]; ok {
if flc.match(p, c) {
return true
}
}
}
for _, v := range fr.CIDR.Supernets(netip.PrefixFrom(p.RemoteAddr, p.RemoteAddr.BitLen())) {
if v.match(p, c) {
return true
}
}
return false
}
func (flc *firewallLocalCIDR) addRule(f *Firewall, localCidr string) error {
if localCidr == "any" {
flc.Any = true
return nil
}
if localCidr == "" {
if len(f.unsafeNetworks) == 0 || f.defaultLocalCIDRAny {
flc.Any = true
return nil
}
for _, network := range f.assignedNetworks {
flc.LocalCIDR.Insert(network)
}
return nil
}
c, err := netip.ParsePrefix(localCidr)
if err != nil {
return err
}
flc.LocalCIDR.Insert(c)
return nil
}
func (flc *firewallLocalCIDR) match(p firewall.Packet, c *cert.CachedCertificate) bool {
if flc == nil {
return false
}
if flc.Any {
return true
}
return flc.LocalCIDR.Contains(p.LocalAddr)
}
type rule struct {
Port string
Code string
Proto string
Host string
Groups []string
Cidr string
LocalCidr string
CAName string
CASha string
}
func convertRule(l *slog.Logger, p any, table string, i int) (rule, error) {
r := rule{}
m, ok := p.(map[string]any)
if !ok {
return r, errors.New("could not parse rule")
}
toString := func(k string, m map[string]any) string {
v, ok := m[k]
if !ok {
return ""
}
return fmt.Sprintf("%v", v)
}
r.Port = toString("port", m)
r.Code = toString("code", m)
r.Proto = toString("proto", m)
r.Host = toString("host", m)
r.Cidr = toString("cidr", m)
r.LocalCidr = toString("local_cidr", m)
r.CAName = toString("ca_name", m)
r.CASha = toString("ca_sha", m)
// Make sure group isn't an array
if v, ok := m["group"].([]any); ok {
if len(v) > 1 {
return r, errors.New("group should contain a single value, an array with more than one entry was provided")
}
l.Warn("group was an array with a single value, converting to simple value",
"table", table,
"rule", i,
)
m["group"] = v[0]
}
singleGroup := toString("group", m)
if rg, ok := m["groups"]; ok {
switch reflect.TypeOf(rg).Kind() {
case reflect.Slice:
v := reflect.ValueOf(rg)
r.Groups = make([]string, v.Len())
for i := 0; i < v.Len(); i++ {
r.Groups[i] = v.Index(i).Interface().(string)
}
case reflect.String:
r.Groups = []string{rg.(string)}
default:
r.Groups = []string{fmt.Sprintf("%v", rg)}
}
}
//flatten group vs groups
if singleGroup != "" {
// Check if we have both groups and group provided in the rule config
if len(r.Groups) > 0 {
return r, fmt.Errorf("only one of group or groups should be defined, both provided")
}
r.Groups = []string{singleGroup}
}
return r, nil
}
// sanity returns an error if the rule would be evaluated in a way that would short-circuit a configured check on a wildcard value
// rules are evaluated as "port AND proto AND (ca_sha OR ca_name) AND (host OR group OR groups OR cidr) AND local_cidr"
func (r *rule) sanity() error {
//port, proto, local_cidr are AND, no need to check here
//ca_sha and ca_name don't have a wildcard value, no need to check here
groupsEmpty := len(r.Groups) == 0
hostEmpty := r.Host == ""
cidrEmpty := r.Cidr == ""
if (groupsEmpty && hostEmpty && cidrEmpty) == true {
return nil //no content!
}
groupsHasAny := slices.Contains(r.Groups, "any")
if groupsHasAny && len(r.Groups) > 1 {
return fmt.Errorf("groups spec [%s] contains the group '\"any\". This rule will ignore the other groups specified", r.Groups)
}
if r.Host == "any" {
if !groupsEmpty {
return fmt.Errorf("groups specified as %s, but host=any will match any host, regardless of groups", r.Groups)
}
if !cidrEmpty {
return fmt.Errorf("cidr specified as %s, but host=any will match any host, regardless of cidr", r.Cidr)
}
}
if groupsHasAny {
if !hostEmpty && r.Host != "any" {
return fmt.Errorf("groups spec [%s] contains the group '\"any\". This rule will ignore the specified host %s", r.Groups, r.Host)
}
if !cidrEmpty {
return fmt.Errorf("groups spec [%s] contains the group '\"any\". This rule will ignore the specified cidr %s", r.Groups, r.Cidr)
}
}
if r.Code != "" {
return fmt.Errorf("code specified as [%s]. Support for 'code' will be dropped in a future release, as it has never been functional", r.Code)
}
//todo alert on cidr-any
return nil
}
func parsePort(s string) (int32, int32, error) {
const notAPort int32 = -2
if s == "any" {
return firewall.PortAny, firewall.PortAny, nil
}
if s == "fragment" {
return firewall.PortFragment, firewall.PortFragment, nil
}
if !strings.Contains(s, `-`) {
rPort, err := parsePortValue("", s)
if err != nil {
return notAPort, notAPort, err
}
return rPort, rPort, nil
}
sPorts := strings.SplitN(s, `-`, 2)
for i := range sPorts {
sPorts[i] = strings.Trim(sPorts[i], " ")
}
if len(sPorts) != 2 || sPorts[0] == "" || sPorts[1] == "" {
return notAPort, notAPort, fmt.Errorf("appears to be a range but could not be parsed; `%s`", s)
}
startPort, err := parsePortValue("beginning range ", sPorts[0])
if err != nil {
return notAPort, notAPort, err
}
endPort, err := parsePortValue("ending range ", sPorts[1])
if err != nil {
return notAPort, notAPort, err
}
if startPort == firewall.PortAny {
endPort = firewall.PortAny
}
return startPort, endPort, nil
}
// parsePortValue accepts a base-10 decimal in [0, 65535] and returns it
// widened to int32. Using strconv.ParseUint with bitSize 16 rejects
// negative input, out-of-range input (>65535), and any non-decimal byte
// by construction, so the int32 widening that follows is provably safe
// and cannot collide with firewall.PortAny (0) or firewall.PortFragment
// (-1) via integer truncation.
//
// prefix is prepended to both error messages so callers can disambiguate
// the single-port path (prefix="") from the range bounds (prefix="beginning
// range " / "ending range "), preserving the historical error strings.
func parsePortValue(prefix, s string) (int32, error) {
n, err := strconv.ParseUint(s, 10, 16)
if err == nil {
return int32(n), nil
}
if errors.Is(err, strconv.ErrRange) {
return 0, fmt.Errorf("%sout of range [0,65535]; `%s`", prefix, s)
}
return 0, fmt.Errorf("%swas not a number; `%s`", prefix, s)
}
+310
View File
@@ -0,0 +1,310 @@
package firewall
import (
"errors"
"fmt"
"log/slog"
"net/netip"
"reflect"
"slices"
"strconv"
"strings"
"github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/iputil"
)
// ruleAdder takes firewall rules, it lets tests see what addRulesFromConfig adds
type ruleAdder interface {
AddRule(incoming bool, proto uint8, startPort int32, endPort int32, groups []string, host string, cidr, localCidr string, caName string, caSha string) error
}
// RulesFromConfig reads the inbound and outbound firewall rules from c and builds them, see RulesBuilder.Build
func RulesFromConfig(l *slog.Logger, c *config.C, vpnNetworks, unsafeNetworks []netip.Prefix) (Rules, error) {
b := NewRulesBuilder(l)
b.defaultLocalCIDRAny = c.GetBool("firewall.default_local_cidr_any", false)
if err := addRulesFromConfig(l, false, c, b); err != nil {
return Rules{}, err
}
if err := addRulesFromConfig(l, true, c, b); err != nil {
return Rules{}, err
}
return b.Build(vpnNetworks, unsafeNetworks), nil
}
func addRulesFromConfig(l *slog.Logger, inbound bool, c *config.C, ra ruleAdder) error {
var table string
if inbound {
table = "firewall.inbound"
} else {
table = "firewall.outbound"
}
r := c.Get(table)
if r == nil {
return nil
}
rs, ok := r.([]any)
if !ok {
return fmt.Errorf("%s failed to parse, should be an array of rules", table)
}
for i, t := range rs {
r, err := yamlToConfigRule(l, t, table, i)
if err != nil {
return fmt.Errorf("%s rule #%v; %s", table, i, err)
}
if r.Code != "" && r.Port != "" {
return fmt.Errorf("%s rule #%v; only one of port or code should be provided", table, i)
}
if r.Host == "" && len(r.Groups) == 0 && r.Cidr == "" && r.LocalCidr == "" && r.CAName == "" && r.CASha == "" {
return fmt.Errorf("%s rule #%v; at least one of host, group, cidr, local_cidr, ca_name, or ca_sha must be provided", table, i)
}
var sPort, errPort string
if r.Code != "" {
errPort = "code"
sPort = r.Code
} else {
errPort = "port"
sPort = r.Port
}
proto, ok := protoByName(r.Proto)
if !ok {
// Any other protocol by number. 0 is reserved for `any`.
n, perr := strconv.ParseUint(r.Proto, 10, 8)
if perr != nil || n == 0 {
return fmt.Errorf("%s rule #%v; proto was not understood; `%s`", table, i, r.Proto)
}
proto = uint8(n)
}
var startPort, endPort int32
if proto == iputil.IPProtocolICMP || proto == iputil.IPProtocolICMPv6 {
// RulesBuilder.AddRule coerces ICMP ports to `any` as well. This copy stays so that an ICMP rule's
// port is warned about rather than parsed, whatever it holds.
startPort = PortAny
endPort = PortAny
if sPort != "" {
l.Warn("ignoring port specification for ICMP firewall rule", "port", sPort)
}
} else {
startPort, endPort, err = parsePort(sPort)
}
if err != nil {
return fmt.Errorf("%s rule #%v; %s %s", table, i, errPort, err)
}
if warning := r.sanity(); warning != nil {
l.Warn("firewall rule sanity check",
"table", table,
"rule", i,
"warning", warning,
)
}
err = ra.AddRule(inbound, proto, startPort, endPort, r.Groups, r.Host, r.Cidr, r.LocalCidr, r.CAName, r.CASha)
if err != nil {
return fmt.Errorf("%s rule #%v; %s", table, i, err)
}
}
return nil
}
type configRule struct {
Port string
Code string
Proto string
Host string
Groups []string
Cidr string
LocalCidr string
CAName string
CASha string
}
func yamlToConfigRule(l *slog.Logger, p any, table string, i int) (configRule, error) {
r := configRule{}
m, ok := p.(map[string]any)
if !ok {
return r, errors.New("could not parse rule")
}
toString := func(k string, m map[string]any) string {
v, ok := m[k]
if !ok {
return ""
}
return fmt.Sprintf("%v", v)
}
r.Port = toString("port", m)
r.Code = toString("code", m)
r.Proto = toString("proto", m)
r.Host = toString("host", m)
r.Cidr = toString("cidr", m)
r.LocalCidr = toString("local_cidr", m)
r.CAName = toString("ca_name", m)
r.CASha = toString("ca_sha", m)
// Make sure group isn't an array
if v, ok := m["group"].([]any); ok {
if len(v) > 1 {
return r, errors.New("group should contain a single value, an array with more than one entry was provided")
}
if len(v) == 0 {
return r, errors.New("group should contain a single value, an empty array was provided")
}
l.Warn("group was an array with a single value, converting to simple value",
"table", table,
"rule", i,
)
m["group"] = v[0]
}
singleGroup := toString("group", m)
// A null value, `groups:` with nothing after it, is the same as no groups
if rg, ok := m["groups"]; ok && rg != nil {
switch reflect.TypeOf(rg).Kind() {
case reflect.Slice:
// Every entry is read as a string, as the other fields are, so `groups: [1, 2]` is the groups "1" and "2"
v := reflect.ValueOf(rg)
r.Groups = make([]string, v.Len())
for i := 0; i < v.Len(); i++ {
r.Groups[i] = fmt.Sprintf("%v", v.Index(i).Interface())
}
case reflect.String:
r.Groups = []string{rg.(string)}
default:
r.Groups = []string{fmt.Sprintf("%v", rg)}
}
}
//flatten group vs groups
if singleGroup != "" {
// Check if we have both groups and group provided in the rule config
if len(r.Groups) > 0 {
return r, fmt.Errorf("only one of group or groups should be defined, both provided")
}
r.Groups = []string{singleGroup}
}
return r, nil
}
// sanity returns an error if the rule would be evaluated in a way that would short-circuit a configured check on a wildcard value
// rules are evaluated as "port AND proto AND (ca_sha OR ca_name) AND (host OR group OR groups OR cidr) AND local_cidr"
func (r *configRule) sanity() error {
//port, proto, local_cidr are AND, no need to check here
//ca_sha and ca_name don't have a wildcard value, no need to check here
groupsEmpty := len(r.Groups) == 0
hostEmpty := r.Host == ""
cidrEmpty := r.Cidr == ""
if (groupsEmpty && hostEmpty && cidrEmpty) == true {
return nil //no content!
}
groupsHasAny := slices.Contains(r.Groups, "any")
if groupsHasAny && len(r.Groups) > 1 {
return fmt.Errorf("groups spec [%s] contains the group '\"any\". This rule will ignore the other groups specified", r.Groups)
}
if r.Host == "any" {
if !groupsEmpty {
return fmt.Errorf("groups specified as %s, but host=any will match any host, regardless of groups", r.Groups)
}
if !cidrEmpty {
return fmt.Errorf("cidr specified as %s, but host=any will match any host, regardless of cidr", r.Cidr)
}
}
if groupsHasAny {
if !hostEmpty && r.Host != "any" {
return fmt.Errorf("groups spec [%s] contains the group '\"any\". This rule will ignore the specified host %s", r.Groups, r.Host)
}
if !cidrEmpty {
return fmt.Errorf("groups spec [%s] contains the group '\"any\". This rule will ignore the specified cidr %s", r.Groups, r.Cidr)
}
}
if r.Code != "" {
return fmt.Errorf("code specified as [%s]. Support for 'code' will be dropped in a future release, as it has never been functional", r.Code)
}
//todo alert on cidr-any
return nil
}
func parsePort(s string) (int32, int32, error) {
const notAPort int32 = -2
if s == "any" {
return PortAny, PortAny, nil
}
if s == "fragment" {
return PortFragment, PortFragment, nil
}
if !strings.Contains(s, `-`) {
rPort, err := parsePortValue("", s)
if err != nil {
return notAPort, notAPort, err
}
return rPort, rPort, nil
}
sPorts := strings.SplitN(s, `-`, 2)
for i := range sPorts {
sPorts[i] = strings.Trim(sPorts[i], " ")
}
if len(sPorts) != 2 || sPorts[0] == "" || sPorts[1] == "" {
return notAPort, notAPort, fmt.Errorf("appears to be a range but could not be parsed; `%s`", s)
}
startPort, err := parsePortValue("beginning range ", sPorts[0])
if err != nil {
return notAPort, notAPort, err
}
endPort, err := parsePortValue("ending range ", sPorts[1])
if err != nil {
return notAPort, notAPort, err
}
if startPort == PortAny {
endPort = PortAny
}
return startPort, endPort, nil
}
// parsePortValue accepts a base-10 decimal in [0, 65535] and returns it
// widened to int32. Using strconv.ParseUint with bitSize 16 rejects
// negative input, out-of-range input (>65535), and any non-decimal byte
// by construction, so the int32 widening that follows is provably safe
// and cannot collide with PortAny (0) or PortFragment
// (-1) via integer truncation.
//
// prefix is prepended to both error messages so callers can disambiguate
// the single-port path (prefix="") from the range bounds (prefix="beginning
// range " / "ending range "), preserving the historical error strings.
func parsePortValue(prefix, s string) (int32, error) {
n, err := strconv.ParseUint(s, 10, 16)
if err == nil {
return int32(n), nil
}
if errors.Is(err, strconv.ErrRange) {
return 0, fmt.Errorf("%sout of range [0,65535]; `%s`", prefix, s)
}
return 0, fmt.Errorf("%swas not a number; `%s`", prefix, s)
}
+516
View File
@@ -0,0 +1,516 @@
package firewall
import (
"bytes"
"errors"
"net/netip"
"testing"
"github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/iputil"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func Test_parsePort(t *testing.T) {
_, _, err := parsePort("")
require.EqualError(t, err, "was not a number; ``")
_, _, err = parsePort(" ")
require.EqualError(t, err, "was not a number; ` `")
_, _, err = parsePort("-")
require.EqualError(t, err, "appears to be a range but could not be parsed; `-`")
_, _, err = parsePort(" - ")
require.EqualError(t, err, "appears to be a range but could not be parsed; ` - `")
_, _, err = parsePort("a-b")
require.EqualError(t, err, "beginning range was not a number; `a`")
_, _, err = parsePort("1-b")
require.EqualError(t, err, "ending range was not a number; `b`")
s, e, err := parsePort(" 1 - 2 ")
assert.Equal(t, int32(1), s)
assert.Equal(t, int32(2), e)
require.NoError(t, err)
s, e, err = parsePort("0-1")
assert.Equal(t, int32(0), s)
assert.Equal(t, int32(0), e)
require.NoError(t, err)
s, e, err = parsePort("9919")
assert.Equal(t, int32(9919), s)
assert.Equal(t, int32(9919), e)
require.NoError(t, err)
s, e, err = parsePort("any")
assert.Equal(t, int32(0), s)
assert.Equal(t, int32(0), e)
require.NoError(t, err)
}
// Test_parsePort_invalid covers inputs that must error. The named bug is
// that int32(strconv.Atoi("4294967296")) truncates to 0 == PortAny,
// silently turning a typo into a match-all-ports rule; the rest are
// representative syntax/range probes.
func Test_parsePort_invalid(t *testing.T) {
tests := []struct {
name string
input string
wantErrContains string
}{
// Numeric overflow (the named bug + boundary).
{"named bug: 2^32 truncates to PortAny", "4294967296", "out of range"},
{"just above max real port", "65536", "out of range"},
// Negatives route through the range branch and hit the empty-half
// guard; included as defense in depth so a future refactor cannot
// accidentally reach the int32 cast.
{"negative", "-1", "could not be parsed"},
// Syntax probes.
{"NUL between digits", "4\x002", "was not a number"},
{"hex notation", "0x10", "was not a number"},
{"scientific notation", "1e3", "was not a number"},
{"leading whitespace", " 42", "was not a number"},
{"fullwidth digits", "42", "was not a number"},
// Range branch.
{"range upper out of range", "1-65536", "ending range out of range"},
{"range lower out of range", "65536-65537", "beginning range out of range"},
{"range with negative upper", "1--1", "ending range was not a number"},
}
for _, tc := range tests {
t.Run(tc.name, func(t *testing.T) {
_, _, err := parsePort(tc.input)
require.Error(t, err, "input %q must error", tc.input)
require.ErrorContains(t, err, tc.wantErrContains)
})
}
}
// Test_parsePort_valid_boundaries locks in success cases at 0, 1, and 65535
// so a future refactor cannot regress the boundaries.
func Test_parsePort_valid_boundaries(t *testing.T) {
tests := []struct {
name string
input string
wantStart int32
wantEnd int32
}{
{"zero is PortAny", "0", 0, 0},
{"min real port", "1", 1, 1},
{"max real port", "65535", 65535, 65535},
{"range zero to max forces end to zero", "0-65535", 0, 0},
{"range max to max", "65535-65535", 65535, 65535},
{"range one to max", "1-65535", 1, 65535},
{"range with whitespace inside", " 1 - 2 ", 1, 2},
}
for _, tc := range tests {
t.Run(tc.name, func(t *testing.T) {
s, e, err := parsePort(tc.input)
require.NoError(t, err)
assert.Equal(t, tc.wantStart, s, "start port")
assert.Equal(t, tc.wantEnd, e, "end port")
})
}
}
func TestAddRulesFromConfig(t *testing.T) {
l := test.NewLogger()
// Test adding tcp rule
conf := config.NewC(test.NewLogger())
mf := &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"port": "1", "proto": "tcp", "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, false, conf, mf))
assert.Equal(t, addRuleCall{incoming: false, proto: iputil.IPProtocolTCP, startPort: 1, endPort: 1, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall)
// Test adding udp rule
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"port": "1", "proto": "udp", "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, false, conf, mf))
assert.Equal(t, addRuleCall{incoming: false, proto: iputil.IPProtocolUDP, startPort: 1, endPort: 1, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall)
// Test adding udplite, dccp, and sctp rules
for name, proto := range map[string]uint8{"udplite": iputil.IPProtocolUDPLite, "dccp": iputil.IPProtocolDCCP, "sctp": iputil.IPProtocolSCTP} {
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"port": "1-2", "proto": name, "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, false, conf, mf), name)
assert.Equal(t, addRuleCall{incoming: false, proto: proto, startPort: 1, endPort: 2, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall, name)
}
// Test adding icmp rule
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"port": "1", "proto": "icmp", "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, false, conf, mf))
assert.Equal(t, addRuleCall{incoming: false, proto: iputil.IPProtocolICMP, startPort: PortAny, endPort: PortAny, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall)
// Test adding icmp rule no port
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"proto": "icmp", "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, false, conf, mf))
assert.Equal(t, addRuleCall{incoming: false, proto: iputil.IPProtocolICMP, startPort: PortAny, endPort: PortAny, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall)
// Test adding rules by protocol number, as yaml would give a string or an int
for _, proto := range []any{"47", 47} {
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"port": "any", "proto": proto, "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, false, conf, mf), proto)
assert.Equal(t, addRuleCall{incoming: false, proto: 47, startPort: PortAny, endPort: PortAny, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall, proto)
}
// Test ICMP and ICMPv6, by name and by number, still ignore ports
for _, tc := range []struct {
given any
want uint8
}{
{"icmp", iputil.IPProtocolICMP},
{"icmpv6", iputil.IPProtocolICMPv6},
{iputil.IPProtocolICMP, iputil.IPProtocolICMP},
{iputil.IPProtocolICMPv6, iputil.IPProtocolICMPv6},
} {
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"port": "1", "proto": tc.given, "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, false, conf, mf), tc.given)
assert.Equal(t, addRuleCall{incoming: false, proto: tc.want, startPort: PortAny, endPort: PortAny, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall, tc.given)
}
// Test protocols that aren't a known name or a number from 1 to 255
for _, proto := range []string{"0", "256", "-1", "gre"} {
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"outbound": []any{map[string]any{"port": "any", "proto": proto, "host": "a"}}}
require.EqualError(t, addRulesFromConfig(l, false, conf, mf), "firewall.outbound rule #0; proto was not understood; `"+proto+"`")
}
// Test adding any rule
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, host: "a", ip: "", localIp: ""}, mf.lastCall)
// Test adding rule with cidr
cidr := netip.MustParsePrefix("10.0.0.0/8")
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "cidr": cidr.String()}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, ip: cidr.String(), localIp: ""}, mf.lastCall)
// Test adding rule with local_cidr
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "local_cidr": cidr.String()}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, ip: "", localIp: cidr.String()}, mf.lastCall)
// Test adding rule with cidr ipv6
cidr6 := netip.MustParsePrefix("fd00::/8")
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "cidr": cidr6.String()}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, ip: cidr6.String(), localIp: ""}, mf.lastCall)
// Test adding rule with any cidr
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "cidr": "any"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, ip: "any", localIp: ""}, mf.lastCall)
// Test adding rule with junk cidr, which RulesBuilder rejects
conf = config.NewC(test.NewLogger())
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "cidr": "junk/junk"}}}
require.EqualError(t, addRulesFromConfig(l, true, conf, NewRulesBuilder(l)), "firewall.inbound rule #0; cidr did not parse; netip.ParsePrefix(\"junk/junk\"): ParseAddr(\"junk\"): unable to parse IP")
// Test adding rule with local_cidr ipv6
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "local_cidr": cidr6.String()}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, ip: "", localIp: cidr6.String()}, mf.lastCall)
// Test adding rule with any local_cidr
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "local_cidr": "any"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, localIp: "any"}, mf.lastCall)
// Test adding rule with junk local_cidr, which RulesBuilder rejects
conf = config.NewC(test.NewLogger())
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "local_cidr": "junk/junk"}}}
require.EqualError(t, addRulesFromConfig(l, true, conf, NewRulesBuilder(l)), "firewall.inbound rule #0; local_cidr did not parse; netip.ParsePrefix(\"junk/junk\"): ParseAddr(\"junk\"): unable to parse IP")
// Test adding rule with ca_sha
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "ca_sha": "12312313123"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, ip: "", localIp: "", caSha: "12312313123"}, mf.lastCall)
// Test adding rule with ca_name
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "ca_name": "root01"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: nil, ip: "", localIp: "", caName: "root01"}, mf.lastCall)
// Test single group
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "group": "a"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: []string{"a"}, ip: "", localIp: ""}, mf.lastCall)
// Test single groups
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "groups": "a"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: []string{"a"}, ip: "", localIp: ""}, mf.lastCall)
// Test multiple AND groups
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "groups": []string{"a", "b"}}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, addRuleCall{incoming: true, proto: ProtoAny, startPort: 1, endPort: 1, groups: []string{"a", "b"}, ip: "", localIp: ""}, mf.lastCall)
// Test group and groups values that used to crash: an empty group array is rejected, groups entries that
// aren't strings are read as strings, and a null groups is no groups
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "group": []any{}}}}
require.EqualError(t, addRulesFromConfig(l, true, conf, mf), "firewall.inbound rule #0; group should contain a single value, an empty array was provided")
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "groups": []any{1, 2}}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Equal(t, []string{"1", "2"}, mf.lastCall.groups)
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "groups": nil, "host": "a"}}}
require.NoError(t, addRulesFromConfig(l, true, conf, mf))
assert.Empty(t, mf.lastCall.groups)
assert.Equal(t, "a", mf.lastCall.host)
// Test Add error
conf = config.NewC(test.NewLogger())
mf = &mockFirewall{}
mf.nextCallReturn = errors.New("test error")
conf.Settings["firewall"] = map[string]any{"inbound": []any{map[string]any{"port": "1", "proto": "any", "host": "a"}}}
require.EqualError(t, addRulesFromConfig(l, true, conf, mf), "firewall.inbound rule #0; test error")
}
// TestRulesFromConfig_defaultLocalCIDR ensures rules without a local_cidr allow only the vpn networks when there are
// unsafe networks, unless default_local_cidr_any is set, and that a rule's own local_cidr is kept either way
func TestRulesFromConfig_defaultLocalCIDR(t *testing.T) {
vpnNetworks := []netip.Prefix{netip.MustParsePrefix("10.1.0.5/16")}
unsafeNetworks := []netip.Prefix{netip.MustParsePrefix("192.168.0.0/24")}
vpnAddr := netip.MustParseAddr("10.1.9.9")
unsafeAddr := netip.MustParseAddr("192.168.0.3")
// ruleOnPort returns the rule on proto whose port clause is exactly port, chosen from the rules that a packet
// on port is checked against.
ruleOnPort := func(t *testing.T, table *Table, proto uint8, port int32) *rule {
t.Helper()
for _, r := range table.rulesAt(proto, port) {
if r.startPort == port {
return r
}
}
require.Failf(t, "no rule found", "proto %d port %d", proto, port)
return nil
}
for _, tc := range []struct {
name string
unsafeNetworks []netip.Prefix
defaultLocalCIDRAny bool
wantUnsafe bool
}{
{"no unsafe networks", nil, false, true},
{"unsafe networks", unsafeNetworks, false, false},
{"unsafe networks and default_local_cidr_any", unsafeNetworks, true, true},
} {
t.Run(tc.name, func(t *testing.T) {
conf := config.NewC(test.NewLogger())
conf.Settings["firewall"] = map[string]any{
"default_local_cidr_any": tc.defaultLocalCIDRAny,
"inbound": []any{
map[string]any{"port": "any", "proto": "any", "host": "any"},
map[string]any{"port": "22", "proto": "tcp", "host": "any"},
map[string]any{"port": "80", "proto": "tcp", "host": "any", "local_cidr": "192.168.0.0/24"},
map[string]any{"port": "443", "proto": "tcp", "host": "any", "local_cidr": "any"},
},
}
rules, err := RulesFromConfig(test.NewLogger(), conf, vpnNetworks, tc.unsafeNetworks)
require.NoError(t, err)
// The proto `any` rule, as indexed by a protocol with rules of its own and as shared by one without.
for _, proto := range []uint8{iputil.IPProtocolTCP, 47} {
r := ruleOnPort(t, rules.In, proto, PortAny)
assert.True(t, r.matchLocal(vpnAddr), "proto %d", proto)
assert.Equal(t, tc.wantUnsafe, r.matchLocal(unsafeAddr), "proto %d", proto)
}
r := ruleOnPort(t, rules.In, iputil.IPProtocolTCP, 22)
assert.True(t, r.matchLocal(vpnAddr))
assert.Equal(t, tc.wantUnsafe, r.matchLocal(unsafeAddr))
r = ruleOnPort(t, rules.In, iputil.IPProtocolTCP, 80)
assert.False(t, r.matchLocal(vpnAddr))
assert.True(t, r.matchLocal(unsafeAddr))
r = ruleOnPort(t, rules.In, iputil.IPProtocolTCP, 443)
assert.True(t, r.matchLocal(vpnAddr))
assert.True(t, r.matchLocal(unsafeAddr))
})
}
}
func Test_convertRule(t *testing.T) {
ob := &bytes.Buffer{}
l := test.NewLoggerWithOutput(ob)
// Ensure group array of 1 is converted and a warning is printed
c := map[string]any{
"group": []any{"group1"},
}
r, err := yamlToConfigRule(l, c, "test", 1)
assert.Contains(t, ob.String(), "group was an array with a single value, converting to simple value")
assert.Contains(t, ob.String(), "table=test")
assert.Contains(t, ob.String(), "rule=1")
require.NoError(t, err)
assert.Equal(t, []string{"group1"}, r.Groups)
// Ensure group array of > 1 is errord
ob.Reset()
c = map[string]any{
"group": []any{"group1", "group2"},
}
r, err = yamlToConfigRule(l, c, "test", 1)
assert.Empty(t, ob.String())
require.Error(t, err, "group should contain a single value, an array with more than one entry was provided")
// Make sure a well formed group is alright
ob.Reset()
c = map[string]any{
"group": "group1",
}
r, err = yamlToConfigRule(l, c, "test", 1)
require.NoError(t, err)
assert.Equal(t, []string{"group1"}, r.Groups)
}
func Test_convertRuleSanity(t *testing.T) {
ob := &bytes.Buffer{}
l := test.NewLoggerWithOutput(ob)
noWarningPlease := []map[string]any{
{"group": "group1"},
{"groups": []any{"group2"}},
{"host": "bob"},
{"cidr": "1.1.1.1/1"},
{"groups": []any{"group2"}, "host": "bob"},
{"cidr": "1.1.1.1/1", "host": "bob"},
{"groups": []any{"group2"}, "cidr": "1.1.1.1/1"},
{"groups": []any{"group2"}, "cidr": "1.1.1.1/1", "host": "bob"},
}
for _, c := range noWarningPlease {
r, err := yamlToConfigRule(l, c, "test", 1)
require.NoError(t, err)
require.NoError(t, r.sanity(), "should not generate a sanity warning, %+v", c)
}
yesWarningPlease := []map[string]any{
{"group": "group1"},
{"groups": []any{"group2"}},
{"cidr": "1.1.1.1/1"},
{"groups": []any{"group2"}, "host": "bob"},
{"cidr": "1.1.1.1/1", "host": "bob"},
{"groups": []any{"group2"}, "cidr": "1.1.1.1/1"},
{"groups": []any{"group2"}, "cidr": "1.1.1.1/1", "host": "bob"},
}
for _, c := range yesWarningPlease {
c["host"] = "any"
r, err := yamlToConfigRule(l, c, "test", 1)
require.NoError(t, err)
err = r.sanity()
require.Error(t, err, "I wanted a warning: %+v", c)
}
//reset the list
yesWarningPlease = []map[string]any{
{"group": "group1"},
{"groups": []any{"group2"}},
{"cidr": "1.1.1.1/1"},
{"groups": []any{"group2"}, "host": "bob"},
{"cidr": "1.1.1.1/1", "host": "bob"},
{"groups": []any{"group2"}, "cidr": "1.1.1.1/1"},
{"groups": []any{"group2"}, "cidr": "1.1.1.1/1", "host": "bob"},
}
for _, c := range yesWarningPlease {
r, err := yamlToConfigRule(l, c, "test", 1)
require.NoError(t, err)
r.Groups = append(r.Groups, "any")
err = r.sanity()
require.Error(t, err, "I wanted a warning: %+v", c)
}
}
type addRuleCall struct {
incoming bool
proto uint8
startPort int32
endPort int32
groups []string
host string
ip string
localIp string
caName string
caSha string
}
type mockFirewall struct {
lastCall addRuleCall
nextCallReturn error
}
func (mf *mockFirewall) AddRule(incoming bool, proto uint8, startPort int32, endPort int32, groups []string, host string, ip, localIp, caName string, caSha string) error {
mf.lastCall = addRuleCall{
incoming: incoming,
proto: proto,
startPort: startPort,
endPort: endPort,
groups: groups,
host: host,
ip: ip,
localIp: localIp,
caName: caName,
caSha: caSha,
}
err := mf.nextCallReturn
mf.nextCallReturn = nil
return err
}
+40 -14
View File
@@ -40,29 +40,55 @@ func (fp *Packet) Copy() *Packet {
} }
func (fp Packet) MarshalJSON() ([]byte, error) { func (fp Packet) MarshalJSON() ([]byte, error) {
var proto string
switch fp.Protocol {
case iputil.IPProtocolTCP:
proto = "tcp"
case iputil.IPProtocolICMP:
proto = "icmp"
case iputil.IPProtocolICMPv6:
proto = "icmpv6"
case iputil.IPProtocolUDP:
proto = "udp"
default:
proto = fmt.Sprintf("unknown %v", fp.Protocol)
}
return json.Marshal(m{ return json.Marshal(m{
"LocalAddr": fp.LocalAddr.String(), "LocalAddr": fp.LocalAddr.String(),
"RemoteAddr": fp.RemoteAddr.String(), "RemoteAddr": fp.RemoteAddr.String(),
"LocalPort": fp.LocalPort, "LocalPort": fp.LocalPort,
"RemotePort": fp.RemotePort, "RemotePort": fp.RemotePort,
"Protocol": proto, "Protocol": protoName(fp.Protocol),
"Fragment": fp.Fragment, "Fragment": fp.Fragment,
}) })
} }
// protoNames are the protocol names that rules accept and packets report, and their numbers. Any other protocol
// is given by number.
var protoNames = []struct {
name string
proto uint8
}{
{"any", ProtoAny},
{"tcp", iputil.IPProtocolTCP},
{"udp", iputil.IPProtocolUDP},
{"udplite", iputil.IPProtocolUDPLite},
{"dccp", iputil.IPProtocolDCCP},
{"sctp", iputil.IPProtocolSCTP},
{"icmp", iputil.IPProtocolICMP},
{"icmpv6", iputil.IPProtocolICMPv6},
}
// protoByName returns the number of a protocol named in protoNames.
func protoByName(name string) (uint8, bool) {
for _, p := range protoNames {
if p.name == name {
return p.proto, true
}
}
return 0, false
}
// protoName returns the name in protoNames of a packet's protocol, or "unknown" and its number. `any` is only
// a rule's protocol, so a packet's protocol 0 is unknown.
func protoName(proto uint8) string {
if proto != ProtoAny {
for _, p := range protoNames {
if p.proto == proto {
return p.name
}
}
}
return fmt.Sprintf("unknown %v", proto)
}
// ParsedPacket is a Packet plus the parse byproducts the RX path reuses // ParsedPacket is a Packet plus the parse byproducts the RX path reuses
type ParsedPacket struct { type ParsedPacket struct {
Packet Packet
+258
View File
@@ -0,0 +1,258 @@
package firewall
import (
"crypto/sha256"
"encoding/hex"
"fmt"
"hash/fnv"
"log/slog"
"math"
"net/netip"
"slices"
"strings"
"github.com/slackhq/nebula/iputil"
)
// RulesBuilder collects the inbound and outbound rules, Build turns them into a Table for each direction
type RulesBuilder struct {
// in and out are each direction's rules in the order they were added
in, out []rule
// defaultLocalCIDRAny makes rules without a local cidr allow any local address, even with unsafe networks
defaultLocalCIDRAny bool
// ruleText describes every rule as it is added, one line each. Build hashes it; see Rules.
ruleText strings.Builder
l *slog.Logger
}
func NewRulesBuilder(l *slog.Logger) *RulesBuilder {
return &RulesBuilder{l: l}
}
// AddRule adds a rule for incoming or outgoing traffic, on proto or on every protocol when proto is ProtoAny.
// The rule is parsed now, so a bad rule is reported when it's added rather than when the Tables are built.
func (b *RulesBuilder) AddRule(incoming bool, proto uint8, startPort int32, endPort int32, groups []string, host string, cidr, localCidr, caName string, caSha string) error {
// ICMPv6 rules are ICMP rules, and a packet of either protocol is checked against both
if proto == iputil.IPProtocolICMPv6 {
proto = iputil.IPProtocolICMP
}
if proto == iputil.IPProtocolICMP {
//ICMP traffic doesn't have ports, so we always coerce to "any", even if a value is provided
if startPort != PortAny {
b.l.Warn("ignoring port specification for ICMP firewall rule", "startPort", startPort)
}
startPort = PortAny
endPort = PortAny
}
r, err := parseRule(proto, startPort, endPort, groups, host, cidr, localCidr, caName, caSha)
if err != nil {
return err
}
if incoming {
b.in = append(b.in, r)
} else {
b.out = append(b.out, r)
}
// The rule hashes are of these lines, changing them changes every hash
fmt.Fprintf(&b.ruleText,
"incoming: %v, proto: %v, startPort: %v, endPort: %v, groups: %v, host: %v, ip: %v, localIp: %v, caName: %v, caSha: %s\n",
incoming, proto, startPort, endPort, groups, host, cidr, localCidr, caName, caSha,
)
direction := "incoming"
if !incoming {
direction = "outgoing"
}
b.l.Info("Firewall rule added",
"firewallRule", m{"direction": direction, "proto": proto, "startPort": startPort, "endPort": endPort, "groups": groups, "host": host, "cidr": cidr, "localCidr": localCidr, "caName": caName, "caSha": caSha},
)
return nil
}
// Build turns the rules into a Table for each direction, and hashes them. Rules without a local cidr allow
// vpnNetworks when there are unsafe networks, otherwise any local address.
func (b *RulesBuilder) Build(vpnNetworks, unsafeNetworks []netip.Prefix) Rules {
var defaultLocal []netip.Prefix
if len(unsafeNetworks) > 0 && !b.defaultLocalCIDRAny {
// Never nil, even when there are no vpn networks, so that the rules allow nothing rather than everything.
defaultLocal = append([]netip.Prefix{}, vpnNetworks...)
}
text := []byte(b.ruleText.String())
sha := sha256.Sum256(text)
h := fnv.New32a()
h.Write(text)
return Rules{
In: buildTable(b.in, defaultLocal),
Out: buildTable(b.out, defaultLocal),
Hash: hex.EncodeToString(sha[:]),
HashFNV: h.Sum32(),
}
}
func parseRule(proto uint8, startPort, endPort int32, groups []string, host, cidr, localCidr, caName, caSha string) (rule, error) {
if startPort > endPort {
return rule{}, fmt.Errorf("start port was lower than end port")
}
if startPort < PortFragment || endPort > math.MaxUint16 {
return rule{}, fmt.Errorf("port range %d-%d is outside of %d-%d", startPort, endPort, PortFragment, math.MaxUint16)
}
// A range that includes port 0 is port `any`.
if startPort <= PortAny && PortAny <= endPort {
startPort, endPort = PortAny, PortAny
}
// A port rule on a protocol without ports could never match. ICMP rules were coerced to `any` by AddRule,
// and a proto `any` rule with a port applies to the protocols that have them.
if startPort > PortAny && proto != ProtoAny && !iputil.HasPorts(proto) {
return rule{}, fmt.Errorf("protocol %d has no ports, port must be any or fragment", proto)
}
r := rule{
proto: proto,
startPort: startPort,
endPort: endPort,
groups: groups,
host: host,
caName: caName,
caSha: caSha,
}
var anyCIDR bool
var err error
r.cidr, anyCIDR, err = parseCIDR(cidr)
if err != nil {
return rule{}, fmt.Errorf("cidr did not parse; %w", err)
}
// A rule with no group, host, or cidr allows any host, as does a rule with `any` for one of them.
r.remoteAny = anyCIDR || host == "any" || slices.Contains(groups, "any") ||
(len(groups) == 0 && host == "" && !r.cidr.IsValid())
var localCIDR netip.Prefix
localCIDR, r.localAny, err = parseCIDR(localCidr)
if err != nil {
return rule{}, fmt.Errorf("local_cidr did not parse; %w", err)
}
if localCIDR.IsValid() {
r.local = []netip.Prefix{localCIDR}
}
return r, nil
}
// buildTable turns one direction's rules into a Table. Rules without a local cidr allow defaultLocal,
// or any local address when defaultLocal is nil.
func buildTable(rules []rule, defaultLocal []netip.Prefix) *Table {
t := &Table{rules: make([]rule, len(rules))}
var hasOwn [256]bool
for id, r := range rules {
t.rules[id] = r.withDefaultLocal(defaultLocal)
if r.proto != ProtoAny {
hasOwn[r.proto] = true
}
}
// A protocol with rules of its own gets an index of those rules plus the proto `any` rules, so a packet
// checks a single index. The other protocols share an index of the proto `any` rules, one for protocols with
// ports and one for protocols without.
anyPorts := newProtoIndex(true, t.rules, ProtoAny)
anyNoPorts := newProtoIndex(false, t.rules, ProtoAny)
for proto := range t.protos {
switch {
case hasOwn[proto]:
t.protos[proto] = newProtoIndex(iputil.HasPorts(uint8(proto)), t.rules, uint8(proto))
case iputil.HasPorts(uint8(proto)):
t.protos[proto] = anyPorts
default:
t.protos[proto] = anyNoPorts
}
}
// ICMPv6 rules are ICMP rules, see AddRule, so ICMPv6 packets are checked against the ICMP index
t.protos[iputil.IPProtocolICMPv6] = t.protos[iputil.IPProtocolICMP]
return t
}
// newProtoIndex builds an index of the proto `any` rules and proto's own rules, for a protocol with or without
// ports, or returns nil when there are none. With ProtoAny, it indexes only the proto `any` rules.
func newProtoIndex(hasPorts bool, rules []rule, proto uint8) *protoIndex {
var pi *protoIndex
for id := range rules {
r := &rules[id]
if r.proto != ProtoAny && r.proto != proto {
continue
}
if pi == nil {
pi = &protoIndex{hasPorts: hasPorts, fragment: newRuleSet(len(rules))}
if hasPorts {
pi.byPort = newRuleSets(math.MaxUint16+1, len(rules))
} else {
pi.packet = newRuleSet(len(rules))
}
}
pi.add(r, id)
}
return pi
}
// add puts id in every set that r's port clause covers.
func (pi *protoIndex) add(r *rule, id int) {
switch r.startPort {
case PortAny:
pi.fragment.add(id)
if pi.hasPorts {
pi.addPorts(0, math.MaxUint16, id)
} else {
pi.packet.add(id)
}
case PortFragment:
pi.fragment.add(id)
default:
// A proto `any` rule with a port is indexed for the protocols that have ports, parseRule rejects a port on
// a protocol without them.
if pi.hasPorts {
pi.addPorts(int(r.startPort), int(r.endPort), id)
}
}
}
// addPorts puts id in the set of every port from lo through hi.
func (pi *protoIndex) addPorts(lo, hi, id int) {
for port := lo; port <= hi; port++ {
pi.byPort.at(port).add(id)
}
}
// parseCIDR parses a rule's cidr, which is empty, `any`, or a prefix
func parseCIDR(s string) (prefix netip.Prefix, isAny bool, err error) {
switch s {
case "":
return netip.Prefix{}, false, nil
case "any":
return netip.Prefix{}, true, nil
}
prefix, err = netip.ParsePrefix(s)
return prefix, false, err
}
// withDefaultLocal returns r allowing defaultLocal, or any local address when defaultLocal is nil,
// unless r has a local cidr of its own
func (r rule) withDefaultLocal(defaultLocal []netip.Prefix) rule {
if !r.localAny && r.local == nil {
r.localAny = defaultLocal == nil
r.local = defaultLocal
}
return r
}
+367
View File
@@ -0,0 +1,367 @@
package firewall
import (
"fmt"
"math"
"net/netip"
"testing"
"time"
"github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/cert_test"
"github.com/slackhq/nebula/iputil"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// rulesAt returns the rules that a packet on proto with the given port is checked against, in id order.
// port may also be PortAny or PortFragment.
func (t *Table) rulesAt(proto uint8, port int32) []*rule {
pi := t.protos[proto]
if pi == nil {
return nil
}
p := &Packet{Protocol: proto, Fragment: port == PortFragment}
if port > 0 {
p.LocalPort = uint16(port)
}
var rules []*rule
for id := range pi.whichMatch(p, true).all() {
rules = append(rules, &t.rules[id])
}
return rules
}
// oneRuleAt returns the single rule that rulesAt finds, and fails the test when there is not exactly one.
func oneRuleAt(t *testing.T, table *Table, proto uint8, key int32) *rule {
t.Helper()
rules := table.rulesAt(proto, key)
require.Len(t, rules, 1)
return rules[0]
}
func TestRulesBuilder_AddRule(t *testing.T) {
l := test.NewLogger()
ti := netip.MustParsePrefix("1.2.3.4/32")
ti6 := netip.MustParsePrefix("fd12::34/128")
rb := NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 1, 1, []string{}, "", "", "", "", ""))
rules := rb.Build(nil, nil)
// An empty rule is any
r := oneRuleAt(t, rules.In, iputil.IPProtocolTCP, 1)
assert.True(t, r.remoteAny)
assert.Empty(t, r.groups)
assert.Empty(t, r.host)
// It applies only to its own port.
assert.Empty(t, rules.In.rulesAt(iputil.IPProtocolTCP, 2))
assert.Empty(t, rules.In.rulesAt(iputil.IPProtocolTCP, PortAny))
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, iputil.IPProtocolUDP, 1, 1, []string{"g1"}, "", "", "", "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.In, iputil.IPProtocolUDP, 1)
assert.False(t, r.remoteAny)
assert.Equal(t, []string{"g1"}, r.groups)
assert.Empty(t, r.host)
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, iputil.IPProtocolICMP, 1, 1, []string{}, "h1", "", "", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolICMPv6, 1, 1, []string{}, "h2", "", "", "", ""))
rules = rb.Build(nil, nil)
// An ICMP rule is port `any` regardless of the port given, and ICMP and ICMPv6 share one set of rules.
for _, proto := range []uint8{iputil.IPProtocolICMP, iputil.IPProtocolICMPv6} {
rs := rules.In.rulesAt(proto, PortAny)
require.Len(t, rs, 2, "proto %d", proto)
assert.False(t, rs[0].remoteAny)
assert.Equal(t, "h1", rs[0].host)
assert.Equal(t, "h2", rs[1].host)
}
assert.Same(t, rules.In.protos[iputil.IPProtocolICMP], rules.In.protos[iputil.IPProtocolICMPv6])
// Any other protocol number gets its own rules
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, 47, 0, 0, []string{"g1"}, "", "", "", "", "")) // GRE
rules = rb.Build(nil, nil)
assert.Equal(t, []string{"g1"}, oneRuleAt(t, rules.In, 47, PortAny).groups)
assert.Nil(t, rules.In.protos[iputil.IPProtocolTCP])
// A proto `any` rule with a port applies to every protocol with ports, and ProtoAny itself has none.
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 1, 1, []string{}, "", ti.String(), "", "", ""))
rules = rb.Build(nil, nil)
assert.Empty(t, rules.Out.rulesAt(ProtoAny, 1))
r = oneRuleAt(t, rules.Out, iputil.IPProtocolTCP, 1)
assert.False(t, r.remoteAny)
assert.Equal(t, ti, r.cidr)
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 1, 1, []string{}, "", ti6.String(), "", "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.Out, iputil.IPProtocolTCP, 1)
assert.False(t, r.remoteAny)
assert.Equal(t, ti6, r.cidr)
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 1, 1, []string{}, "", "", ti.String(), "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.Out, iputil.IPProtocolTCP, 1)
assert.True(t, r.remoteAny)
assert.False(t, r.localAny)
assert.Equal(t, []netip.Prefix{ti}, r.local)
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 1, 1, []string{}, "", "", ti6.String(), "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.Out, iputil.IPProtocolTCP, 1)
assert.True(t, r.remoteAny)
assert.False(t, r.localAny)
assert.Equal(t, []netip.Prefix{ti6}, r.local)
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, iputil.IPProtocolUDP, 1, 1, []string{"g1"}, "", "", "", "ca-name", ""))
rules = rb.Build(nil, nil)
assert.Equal(t, "ca-name", oneRuleAt(t, rules.In, iputil.IPProtocolUDP, 1).caName)
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, iputil.IPProtocolUDP, 1, 1, []string{"g1"}, "", "", "", "", "ca-sha"))
rules = rb.Build(nil, nil)
assert.Equal(t, "ca-sha", oneRuleAt(t, rules.In, iputil.IPProtocolUDP, 1).caSha)
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 0, 0, []string{}, "any", "", "", "", ""))
rules = rb.Build(nil, nil)
assert.True(t, oneRuleAt(t, rules.Out, ProtoAny, 0).remoteAny)
// 0.0.0.0/0 allows any IPv4 address, but no IPv6 address.
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 0, 0, []string{}, "", "0.0.0.0/0", "", "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.Out, ProtoAny, 0)
assert.False(t, r.remoteAny)
assert.True(t, r.matchRemote(netip.MustParseAddr("1.1.1.1"), nil))
assert.False(t, r.matchRemote(netip.MustParseAddr("9::9"), nil))
// ::/0 allows any IPv6 address, but no IPv4 address.
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 0, 0, []string{}, "", "::/0", "", "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.Out, ProtoAny, 0)
assert.False(t, r.remoteAny)
assert.True(t, r.matchRemote(netip.MustParseAddr("9::9"), nil))
assert.False(t, r.matchRemote(netip.MustParseAddr("1.1.1.1"), nil))
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 0, 0, []string{}, "", "any", "", "", ""))
rules = rb.Build(nil, nil)
assert.True(t, oneRuleAt(t, rules.Out, ProtoAny, 0).remoteAny)
// A local cidr is likewise limited to its own address family.
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 0, 0, []string{}, "", "", "0.0.0.0/0", "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.Out, ProtoAny, 0)
assert.False(t, r.localAny)
assert.True(t, r.matchLocal(netip.MustParseAddr("1.1.1.1")))
assert.False(t, r.matchLocal(netip.MustParseAddr("9::9")))
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 0, 0, []string{}, "", "", "::/0", "", ""))
rules = rb.Build(nil, nil)
r = oneRuleAt(t, rules.Out, ProtoAny, 0)
assert.False(t, r.localAny)
assert.True(t, r.matchLocal(netip.MustParseAddr("9::9")))
assert.False(t, r.matchLocal(netip.MustParseAddr("1.1.1.1")))
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(false, ProtoAny, 0, 0, []string{}, "", "", "any", "", ""))
rules = rb.Build(nil, nil)
assert.True(t, oneRuleAt(t, rules.Out, ProtoAny, 0).localAny)
// Every protocol number is accepted
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, math.MaxUint8, 0, 0, []string{}, "", "", "", "", ""))
rules = rb.Build(nil, nil)
assert.True(t, oneRuleAt(t, rules.In, math.MaxUint8, PortAny).remoteAny)
// Test error conditions, a bad rule is reported when it's added
rb = NewRulesBuilder(l)
require.Error(t, rb.AddRule(true, ProtoAny, 10, 0, []string{}, "", "", "", "", ""))
require.Error(t, rb.AddRule(true, ProtoAny, PortFragment-1, 0, []string{}, "", "", "", "", ""))
require.Error(t, rb.AddRule(true, ProtoAny, 1, math.MaxUint16+1, []string{}, "", "", "", "", ""))
require.Error(t, rb.AddRule(true, ProtoAny, 0, 0, []string{}, "", "junk", "", "", ""))
require.Error(t, rb.AddRule(true, ProtoAny, 0, 0, []string{}, "", "", "junk", "", ""))
// ICMP and ICMPv6 share one copy of each proto `any` rule
rb = NewRulesBuilder(l)
require.NoError(t, rb.AddRule(true, ProtoAny, PortAny, PortAny, []string{"default-group"}, "", "", "", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolICMPv6, PortAny, PortAny, []string{"nope"}, "", "", "", "", ""))
rules = rb.Build(nil, nil)
assert.Len(t, rules.In.rules, 2)
assert.Len(t, rules.In.rulesAt(iputil.IPProtocolICMP, PortAny), 2)
assert.Len(t, rules.In.rulesAt(iputil.IPProtocolICMPv6, PortAny), 2)
// No rules allow nothing
rules = NewRulesBuilder(l).Build(nil, nil)
for proto := range rules.In.protos {
assert.Nil(t, rules.In.protos[proto], "proto %d", proto)
assert.Nil(t, rules.Out.protos[proto], "proto %d", proto)
}
}
// TestRulesBuilder_protoIndex checks which packets each kind of port clause covers, and which protocols
// share an index.
func TestRulesBuilder_protoIndex(t *testing.T) {
rb := NewRulesBuilder(test.NewLogger())
require.NoError(t, rb.AddRule(true, ProtoAny, PortAny, PortAny, nil, "any", "", "", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 22, 22, nil, "port", "", "", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 1024, math.MaxUint16, nil, "range", "", "", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, PortFragment, PortFragment, nil, "fragment", "", "", "", ""))
// A range that includes port 0 is port `any`.
require.NoError(t, rb.AddRule(true, iputil.IPProtocolUDP, 0, 100, nil, "zero", "", "", "", ""))
// GRE has no ports, so a port rule on it is rejected rather than never applying, and `any` and `fragment` apply
require.Error(t, rb.AddRule(true, 47, 22, 22, nil, "gre", "", "", "", ""))
require.NoError(t, rb.AddRule(true, 47, PortFragment, PortFragment, nil, "gre", "", "", "", ""))
rules := rb.Build(nil, nil)
hosts := func(proto uint8, key int32) []string {
var hosts []string
for _, r := range rules.In.rulesAt(proto, key) {
hosts = append(hosts, r.host)
}
return hosts
}
assert.Equal(t, []string{"any", "fragment"}, hosts(iputil.IPProtocolTCP, PortFragment))
assert.Equal(t, []string{"any"}, hosts(iputil.IPProtocolTCP, PortAny))
assert.Equal(t, []string{"any"}, hosts(iputil.IPProtocolTCP, 21))
assert.Equal(t, []string{"any", "port"}, hosts(iputil.IPProtocolTCP, 22))
assert.Equal(t, []string{"any"}, hosts(iputil.IPProtocolTCP, 23))
assert.Equal(t, []string{"any"}, hosts(iputil.IPProtocolTCP, 1023))
assert.Equal(t, []string{"any", "range"}, hosts(iputil.IPProtocolTCP, 1024))
assert.Equal(t, []string{"any", "range"}, hosts(iputil.IPProtocolTCP, math.MaxUint16))
assert.Equal(t, []string{"any", "zero"}, hosts(iputil.IPProtocolUDP, PortFragment))
assert.Equal(t, []string{"any", "zero"}, hosts(iputil.IPProtocolUDP, PortAny))
assert.Equal(t, []string{"any", "zero"}, hosts(iputil.IPProtocolUDP, math.MaxUint16))
// A protocol without ports sees only the port `any` rules, and the port `fragment` rules for a fragment.
assert.False(t, rules.In.protos[47].hasPorts)
assert.Equal(t, []string{"any"}, hosts(47, PortAny))
assert.Equal(t, []string{"any"}, hosts(47, 22))
assert.Equal(t, []string{"any", "gre"}, hosts(47, PortFragment))
assert.Equal(t, []string{"any"}, hosts(iputil.IPProtocolICMP, PortFragment))
// Protocols without rules of their own share an index of the proto `any` rules: one for protocols with
// ports and one for protocols without.
assert.True(t, rules.In.protos[iputil.IPProtocolSCTP].hasPorts)
assert.Equal(t, []string{"any"}, hosts(iputil.IPProtocolSCTP, 5000))
assert.Same(t, rules.In.protos[iputil.IPProtocolSCTP], rules.In.protos[iputil.IPProtocolDCCP])
assert.NotSame(t, rules.In.protos[iputil.IPProtocolSCTP], rules.In.protos[iputil.IPProtocolTCP])
assert.Equal(t, []string{"any"}, hosts(50, PortAny))
assert.Same(t, rules.In.protos[ProtoAny], rules.In.protos[50])
assert.Same(t, rules.In.protos[ProtoAny], rules.In.protos[iputil.IPProtocolICMP])
assert.NotSame(t, rules.In.protos[ProtoAny], rules.In.protos[iputil.IPProtocolSCTP])
assert.Nil(t, rules.Out.protos[iputil.IPProtocolTCP])
}
// TestRulesBuilder_manyRules checks that rule ids beyond the first byte of a rule set are indexed correctly.
func TestRulesBuilder_manyRules(t *testing.T) {
rb := NewRulesBuilder(test.NewLogger())
const n = 130
for i := range n {
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, int32(i+1), int32(i+1), nil, fmt.Sprintf("h%d", i), "", "", "", ""))
}
require.NoError(t, rb.AddRule(true, ProtoAny, PortAny, PortAny, nil, "any", "", "", "", ""))
rules := rb.Build(nil, nil)
assert.Equal(t, ruleSetLen(n+1), rules.In.protos[iputil.IPProtocolTCP].byPort.setLen)
// Ids are the order the rules were added in, so the proto `any` rule comes last
for i := range n {
rs := rules.In.rulesAt(iputil.IPProtocolTCP, int32(i+1))
require.Len(t, rs, 2, "port %d", i+1)
assert.Equal(t, fmt.Sprintf("h%d", i), rs[0].host)
assert.Equal(t, "any", rs[1].host)
}
assert.Len(t, rules.In.rulesAt(iputil.IPProtocolTCP, n+1), 1)
}
// TestRulesBuilder_defaultLocalCIDR checks which local addresses rules allow when there are unsafe networks.
// A rule without a local cidr allows only the vpn networks, and a rule with one allows only that cidr.
func TestRulesBuilder_defaultLocalCIDR(t *testing.T) {
vpnNetworks := []netip.Prefix{
netip.MustParsePrefix("10.1.0.5/16"),
netip.MustParsePrefix("10.2.0.5/16"),
netip.MustParsePrefix("fd00:1::5/64"),
}
unsafeNetworks := []netip.Prefix{netip.MustParsePrefix("192.168.0.0/24")}
vpn1 := netip.MustParseAddr("10.1.9.9")
vpn2 := netip.MustParseAddr("10.2.9.9")
vpn6 := netip.MustParseAddr("fd00:1::9")
h1Extra := netip.MustParseAddr("10.3.0.1")
h1Extra6 := netip.MustParseAddr("fd00:2::1")
h2Own := netip.MustParseAddr("10.4.0.1")
rb := NewRulesBuilder(test.NewLogger())
// h1 allows the vpn networks by default on ports 1 and 2, and 10.3.0.0/16 and fd00:2::/64 as well on port 1
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 1, 2, nil, "h1", "", "", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 1, 1, nil, "h1", "", "10.3.0.0/16", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 1, 1, nil, "h1", "", "fd00:2::/64", "", ""))
// h2 allows 10.4.0.0/16 on ports 1 and 2, and the vpn networks by default as well on port 1
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 1, 2, nil, "h2", "", "10.4.0.0/16", "", ""))
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 1, 1, nil, "h2", "", "", "", ""))
// h3 only has the default
require.NoError(t, rb.AddRule(true, iputil.IPProtocolTCP, 3, 3, nil, "h3", "", "", "", ""))
rules := rb.Build(vpnNetworks, unsafeNetworks)
ca, _, caKey, _ := cert_test.NewTestCaCert(cert.Version2, cert.Curve_CURVE25519, time.Time{}, time.Time{}, nil, nil, nil)
caPool := cert.NewCAPool()
require.NoError(t, caPool.AddCA(ca))
certs := map[string]*cert.CachedCertificate{}
for _, name := range []string{"h1", "h2", "h3"} {
crt, _, _, _ := cert_test.NewTestCert(cert.Version2, cert.Curve_CURVE25519, ca, caKey, name, time.Time{}, time.Time{}, []netip.Prefix{netip.MustParsePrefix("10.1.0.9/16")}, nil, nil)
c, err := caPool.VerifyCertificate(time.Now(), crt)
require.NoError(t, err)
certs[name] = c
}
for _, tc := range []struct {
port uint16
host string
addr netip.Addr
want bool
}{
{1, "h1", vpn1, true},
{1, "h1", vpn2, true},
{1, "h1", vpn6, true},
{1, "h1", h1Extra, true},
{1, "h1", h1Extra6, true},
{1, "h1", h2Own, false},
{2, "h1", vpn1, true},
{2, "h1", h1Extra, false},
{2, "h1", h1Extra6, false},
{1, "h2", h2Own, true},
{1, "h2", vpn1, true},
{1, "h2", vpn2, true},
{1, "h2", vpn6, true},
{1, "h2", h1Extra, false},
{2, "h2", h2Own, true},
{2, "h2", vpn1, false},
{2, "h2", vpn6, false},
{3, "h3", vpn1, true},
{3, "h3", vpn2, true},
{3, "h3", vpn6, true},
{3, "h3", h1Extra, false},
{3, "h3", h1Extra6, false},
{3, "h3", h2Own, false},
// Only h3 is allowed on port 3.
{3, "h1", vpn1, false},
} {
p := &Packet{Protocol: iputil.IPProtocolTCP, LocalPort: tc.port, LocalAddr: tc.addr, RemoteAddr: netip.MustParseAddr("10.1.0.9")}
assert.Equal(t, tc.want, rules.In.Match(p, true, certs[tc.host], caPool), "port %d, %s, %s", tc.port, tc.host, tc.addr)
}
}
+60
View File
@@ -0,0 +1,60 @@
package firewall
import (
"iter"
"math/bits"
)
// ruleSet is a set of rule ids, stored as a bitset. s[i] holds ids i*8 through i*8+7, one per bit.
// Bytes rather than words keep a set to the size the rule count needs, which matters when there is a set for
// every port number.
type ruleSet []uint8
// ruleSetLen returns the length of a ruleSet that can hold ids below n.
func ruleSetLen(n int) int {
return (n + 7) / 8
}
// newRuleSet returns an empty set that can hold ids below idLimit.
func newRuleSet(idLimit int) ruleSet {
return make(ruleSet, ruleSetLen(idLimit))
}
// add puts id in s.
func (s ruleSet) add(id int) {
s[id/8] |= 1 << (id % 8)
}
// ruleSets holds several ruleSet objects of the same length in a single allocation.
type ruleSets struct {
// setLen is the number of bytes needed to represent one set
setLen int
// bits holds the sets in order. Set i starts at i*setLen.
bits []uint8
}
// newRuleSets returns n empty sets, each of which can hold ids below idLimit.
func newRuleSets(n, idLimit int) ruleSets {
setLen := ruleSetLen(idLimit)
return ruleSets{setLen: setLen, bits: make([]uint8, n*setLen)}
}
// at returns set i.
func (s ruleSets) at(i int) ruleSet {
return s.bits[i*s.setLen : (i+1)*s.setLen]
}
// all returns an iterator over the ids in s, in ascending order.
func (s ruleSet) all() iter.Seq[int] {
return func(yield func(int) bool) {
for i, ids := range s {
for ids != 0 {
id := i*8 + bits.TrailingZeros8(ids)
ids &= ids - 1 // Clear the lowest set bit, which is id's.
if !yield(id) {
return
}
}
}
}
}
+189
View File
@@ -0,0 +1,189 @@
package firewall
import (
"net/netip"
"github.com/slackhq/nebula/cert"
)
// Rules are the Tables built for each direction, with hashes of the rules they were built from
type Rules struct {
In, Out *Table
// Hash is a sha256 of the rules, HashFNV is an FNV-1a of them for use as a metric value.
// Rule sets with the same rules have the same hashes.
Hash string
HashFNV uint32
}
// Table holds the rules for one direction. A packet is allowed when every clause of a single rule allows it:
// proto AND port AND (CA SHA or CA name) AND local CIDR AND (group OR groups OR name OR remote CIDR).
// A Table does not change once built; see RulesBuilder.
type Table struct {
// rules holds every rule for the direction, in the order they were added. A rule's index in this slice is
// its id in the ruleSets.
rules []rule
// protos holds the index for each IP protocol number, so a packet is checked against a single index.
// A protocol with rules of its own has an index that also includes every proto `any` rule. The remaining
// protocols share an index of only the proto `any` rules: one for protocols with ports and one for
// protocols without; see hasPorts. An entry is nil when the protocol has no rules at all.
// ICMP and ICMPv6 share one index.
protos [256]*protoIndex
}
// protoIndex holds the rules that can apply to one protocol, in a set for each kind of packet. Rule ids are
// indexes into the direction's rules, so every set is sized by the direction's rule count, not the protocol's:
// byPort costs 64 KB per 8 rules in the direction.
type protoIndex struct {
// hasPorts reports whether the protocol has ports; see iputil.HasPorts.
hasPorts bool
// byPort has a set for every port number of a protocol with ports. Port 0 is covered only by port `any`
// rules, since a range starts at 1.
byPort ruleSets
// packet holds the rules for a packet of a protocol without ports: the port `any` rules.
packet ruleSet
// fragment holds the rules for a fragment of any protocol: the port `any` and port `fragment` rules.
fragment ruleSet
}
// whichMatch returns the set of rules whose proto and port clauses allow p.
func (pi *protoIndex) whichMatch(p *Packet, incoming bool) ruleSet {
switch {
case p.Fragment:
// The ports of a fragmented packet are in its first fragment, and this is a later one.
return pi.fragment
case !pi.hasPorts:
// The port fields are zero, or for ICMP the identifier, which is only for connection tracking.
return pi.packet
case incoming:
return pi.byPort.at(int(p.LocalPort))
default:
return pi.byPort.at(int(p.RemotePort))
}
}
// rule is a parsed rule. A packet is allowed by the rule when its port, certificate, and address clauses all pass.
type rule struct {
// proto is the IP protocol number the rule applies to, or ProtoAny for every protocol. ICMPv6 rules are kept
// as ICMP. The protoIndex checks it, so match does not.
proto uint8
// startPort and endPort are the port clause. A startPort of PortAny covers every port, every packet of a
// protocol without ports, and every fragment. PortFragment covers only fragments. A range covers the ports
// in it, so it never applies to a protocol without ports.
startPort, endPort int32
// caSha and caName pass a certificate issued by either CA. When neither is set, any CA passes.
caSha, caName string
// remoteAny makes host, groups, and cidr irrelevant. Otherwise, the peer passes when any one of them matches.
remoteAny bool
host string
groups []string
cidr netip.Prefix // Invalid when the rule has no cidr.
// localAny allows any local address. Otherwise, local must contain the address. Neither is set for a rule
// without a local cidr until the Table is built and the rule gets the default; see withDefaultLocal.
localAny bool
local []netip.Prefix
}
// Match reports whether a rule for the direction p is going allows it
func (r *Rules) Match(p *Packet, incoming bool, c *cert.CachedCertificate, caPool *cert.CAPool) bool {
t := r.Out
if incoming {
t = r.In
}
return t.Match(p, incoming, c, caPool)
}
// Match reports whether a rule allows p
func (t *Table) Match(p *Packet, incoming bool, c *cert.CachedCertificate, caPool *cert.CAPool) bool {
proto := t.protos[p.Protocol]
if proto == nil {
// The protocol has no rules of its own, and there are no proto `any` rules.
return false
}
// The index yields every rule whose proto and port clauses allow p.
// The packet is allowed when the remaining clauses of any of those rules allow it as well.
for id := range proto.whichMatch(p, incoming).all() {
if t.rules[id].match(p, c, caPool) {
return true
}
}
return false
}
// match reports whether r allows p from the peer with certificate c. The port clause is not checked here; the
// protoIndex has already done so.
func (r *rule) match(p *Packet, c *cert.CachedCertificate, caPool *cert.CAPool) bool {
return r.matchLocal(p.LocalAddr) && r.matchRemote(p.RemoteAddr, c) && r.matchCA(c, caPool)
}
// matchLocal reports whether r allows the local address addr.
func (r *rule) matchLocal(addr netip.Addr) bool {
if r.localAny {
return true
}
for _, prefix := range r.local {
if prefix.Contains(addr) {
return true
}
}
return false
}
// matchRemote reports whether r allows the peer with certificate c, sending from addr. The cidr is checked with
// Prefix.Contains, so 0.0.0.0/0 allows any IPv4 address and ::/0 allows any IPv6 address; neither allows the
// other family.
func (r *rule) matchRemote(addr netip.Addr, c *cert.CachedCertificate) bool {
if r.remoteAny {
return true
}
if r.cidr.IsValid() && r.cidr.Contains(addr) {
return true
}
if r.host != "" && r.host == c.Certificate.Name() {
return true
}
// Every group must be present.
if len(r.groups) == 0 {
return false
}
for _, g := range r.groups {
if _, ok := c.InvertedGroups[g]; !ok {
return false
}
}
return true
}
// matchCA reports whether c was issued by a CA that r allows.
func (r *rule) matchCA(c *cert.CachedCertificate, caPool *cert.CAPool) bool {
if r.caSha == "" && r.caName == "" {
return true
}
if r.caSha != "" && r.caSha == c.Certificate.Issuer() {
return true
}
if r.caName != "" {
s, err := caPool.GetCAForCert(c.Certificate)
if err != nil {
return false
}
return r.caName == s.Certificate.Name()
}
return false
}
+333 -574
View File
File diff suppressed because it is too large. Load diff
+4 -5
View File
@@ -1,13 +1,13 @@
module github.com/slackhq/nebula module github.com/slackhq/nebula
go 1.27.2 go 1.26.0
require ( require (
dario.cat/mergo v1.0.2 dario.cat/mergo v1.0.2
filippo.io/bigmod v0.1.0 filippo.io/bigmod v0.1.0
github.com/DefinedNet/netlink v1.3.2-dn.1 github.com/DefinedNet/netlink v1.3.2-dn.1
github.com/Microsoft/go-winio v0.6.2
github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be
github.com/armon/go-radix v1.0.0
github.com/flynn/noise v1.1.0 github.com/flynn/noise v1.1.0
github.com/gaissmai/bart v0.29.0 github.com/gaissmai/bart v0.29.0
github.com/gogo/protobuf v1.3.2 github.com/gogo/protobuf v1.3.2
@@ -33,7 +33,7 @@ require (
golang.zx2c4.com/wireguard/windows v1.0.1 golang.zx2c4.com/wireguard/windows v1.0.1
google.golang.org/protobuf v1.36.12 google.golang.org/protobuf v1.36.12
gopkg.in/yaml.v3 v3.0.1 gopkg.in/yaml.v3 v3.0.1
gvisor.dev/gvisor v0.0.0-20261014193016-04d31b385980 gvisor.dev/gvisor v0.0.0-20240423190808-9d7a357edefe
) )
require ( require (
@@ -45,6 +45,5 @@ require (
github.com/prometheus/common v0.70.1 // indirect github.com/prometheus/common v0.70.1 // indirect
github.com/prometheus/procfs v0.21.1 // indirect github.com/prometheus/procfs v0.21.1 // indirect
github.com/vishvananda/netns v0.0.5 // indirect github.com/vishvananda/netns v0.0.5 // indirect
golang.org/x/exp v0.0.0-20250711185948-6ae5c78190dc // indirect golang.org/x/time v0.5.0 // indirect
golang.org/x/time v0.15.0 // indirect
) )
+6 -8
View File
@@ -5,8 +5,6 @@ filippo.io/bigmod v0.1.0 h1:UNzDk7y9ADKST+axd9skUpBQeW7fG2KrTZyOE4uGQy8=
filippo.io/bigmod v0.1.0/go.mod h1:OjOXDNlClLblvXdwgFFOQFJEocLhhtai8vGLy0JCZlI= filippo.io/bigmod v0.1.0/go.mod h1:OjOXDNlClLblvXdwgFFOQFJEocLhhtai8vGLy0JCZlI=
github.com/DefinedNet/netlink v1.3.2-dn.1 h1:Md2PsUE1Xuy1Zxs/Bm6356n1kt5bq2HspXlPJ6pEspg= github.com/DefinedNet/netlink v1.3.2-dn.1 h1:Md2PsUE1Xuy1Zxs/Bm6356n1kt5bq2HspXlPJ6pEspg=
github.com/DefinedNet/netlink v1.3.2-dn.1/go.mod h1:o4YSn8Jw7zoDcYXDmV77TRd4eS5sSLuE+vu2zAN67C8= github.com/DefinedNet/netlink v1.3.2-dn.1/go.mod h1:o4YSn8Jw7zoDcYXDmV77TRd4eS5sSLuE+vu2zAN67C8=
github.com/Microsoft/go-winio v0.6.2 h1:F2VQgta7ecxGYO8k3ZZz3RS8fVIXVxONVUPlNERoyfY=
github.com/Microsoft/go-winio v0.6.2/go.mod h1:yd8OoFMLzJbo9gZq8j5qaps8bJ9aShtEA8Ipt1oGCvU=
github.com/alecthomas/template v0.0.0-20160405071501-a0175ee3bccc/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc= github.com/alecthomas/template v0.0.0-20160405071501-a0175ee3bccc/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc=
github.com/alecthomas/template v0.0.0-20190718012654-fb15b899a751/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc= github.com/alecthomas/template v0.0.0-20190718012654-fb15b899a751/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc=
github.com/alecthomas/units v0.0.0-20151022065526-2efee857e7cf/go.mod h1:ybxpYRFXyAe+OPACYpWeL0wqObRcbAqCMya13uyzqw0= github.com/alecthomas/units v0.0.0-20151022065526-2efee857e7cf/go.mod h1:ybxpYRFXyAe+OPACYpWeL0wqObRcbAqCMya13uyzqw0=
@@ -14,6 +12,8 @@ github.com/alecthomas/units v0.0.0-20190717042225-c3de453c63f4/go.mod h1:ybxpYRF
github.com/alecthomas/units v0.0.0-20190924025748-f65c72e2690d/go.mod h1:rBZYJk541a8SKzHPHnH3zbiI+7dagKZ0cgpgrD7Fyho= github.com/alecthomas/units v0.0.0-20190924025748-f65c72e2690d/go.mod h1:rBZYJk541a8SKzHPHnH3zbiI+7dagKZ0cgpgrD7Fyho=
github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be h1:9AeTilPcZAjCFIImctFaOjnTIavg87rW78vTPkQqLI8= github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be h1:9AeTilPcZAjCFIImctFaOjnTIavg87rW78vTPkQqLI8=
github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be/go.mod h1:ySMOLuWl6zY27l47sB3qLNK6tF2fkHG55UZxx8oIVo4= github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be/go.mod h1:ySMOLuWl6zY27l47sB3qLNK6tF2fkHG55UZxx8oIVo4=
github.com/armon/go-radix v1.0.0 h1:F4z6KzEeeQIMeLFa97iZU6vupzoecKdU5TX24SNppXI=
github.com/armon/go-radix v1.0.0/go.mod h1:ufUuZ+zHj4x4TnLV4JWEpy2hxWSpsRywHrMgIH9cCH8=
github.com/beorn7/perks v0.0.0-20180321164747-3a771d992973/go.mod h1:Dwedo/Wpr24TaqPxmxbtue+5NUziq4I4S80YR8gNf3Q= github.com/beorn7/perks v0.0.0-20180321164747-3a771d992973/go.mod h1:Dwedo/Wpr24TaqPxmxbtue+5NUziq4I4S80YR8gNf3Q=
github.com/beorn7/perks v1.0.0/go.mod h1:KWe93zE9D1o94FZ5RNwFwVgaQK1VOXiVxmqh+CedLV8= github.com/beorn7/perks v1.0.0/go.mod h1:KWe93zE9D1o94FZ5RNwFwVgaQK1VOXiVxmqh+CedLV8=
github.com/beorn7/perks v1.0.1 h1:VlbKKnNfV8bJzeqoa4cOKqO6bYr3WgKZxO8Z16+hsOM= github.com/beorn7/perks v1.0.1 h1:VlbKKnNfV8bJzeqoa4cOKqO6bYr3WgKZxO8Z16+hsOM=
@@ -157,8 +157,6 @@ golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9/go.mod h1:LzIPMQfyMNhhGPh
golang.org/x/crypto v0.0.0-20210322153248-0c34fe9e7dc2/go.mod h1:T9bdIzuCu7OtxOm1hfPfRQxPLYneinmdGuTeoZ9dtd4= golang.org/x/crypto v0.0.0-20210322153248-0c34fe9e7dc2/go.mod h1:T9bdIzuCu7OtxOm1hfPfRQxPLYneinmdGuTeoZ9dtd4=
golang.org/x/crypto v0.57.0 h1:3ZVCjf8Ggz7zneR/EHRVx68Ctf+2pmIMP2UFhh9cC6M= golang.org/x/crypto v0.57.0 h1:3ZVCjf8Ggz7zneR/EHRVx68Ctf+2pmIMP2UFhh9cC6M=
golang.org/x/crypto v0.57.0/go.mod h1:Fdz0i5U6CoizGwLda9DttjSk6qlZo25zYNtR+ycvuZA= golang.org/x/crypto v0.57.0/go.mod h1:Fdz0i5U6CoizGwLda9DttjSk6qlZo25zYNtR+ycvuZA=
golang.org/x/exp v0.0.0-20250711185948-6ae5c78190dc h1:TS73t7x3KarrNd5qAipmspBDS1rkMcgVG/fS1aRb4Rc=
golang.org/x/exp v0.0.0-20250711185948-6ae5c78190dc/go.mod h1:A+z0yzpGtvnG90cToK5n2tu8UJVP2XUATh+r+sfOOOc=
golang.org/x/lint v0.0.0-20200302205851-738671d3881b/go.mod h1:3xt1FjdF8hUf6vQPIChWIBhFzV8gjjsPE/fR3IyQdNY= golang.org/x/lint v0.0.0-20200302205851-738671d3881b/go.mod h1:3xt1FjdF8hUf6vQPIChWIBhFzV8gjjsPE/fR3IyQdNY=
golang.org/x/mod v0.1.1-0.20191105210325-c90efee705ee/go.mod h1:QqPTAvyqsEbceGzBzNggFXnrqF1CaUcvgkdR5Ot7KZg= golang.org/x/mod v0.1.1-0.20191105210325-c90efee705ee/go.mod h1:QqPTAvyqsEbceGzBzNggFXnrqF1CaUcvgkdR5Ot7KZg=
golang.org/x/mod v0.2.0/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA= golang.org/x/mod v0.2.0/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA=
@@ -205,8 +203,8 @@ golang.org/x/term v0.46.0/go.mod h1:+K02xbkittuwc0Am4abfA3Fc+XRGXkvBXNO88NCXPoc=
golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ= golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
golang.org/x/text v0.3.2/go.mod h1:bEr9sfX3Q8Zfm5fL9x+3itogRgK3+ptLWKqgva+5dAk= golang.org/x/text v0.3.2/go.mod h1:bEr9sfX3Q8Zfm5fL9x+3itogRgK3+ptLWKqgva+5dAk=
golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ= golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
golang.org/x/time v0.15.0 h1:bbrp8t3bGUeFOx08pvsMYRTCVSMk89u4tKbNOZbp88U= golang.org/x/time v0.5.0 h1:o7cqy6amK/52YcAKIPlM3a+Fpj35zvRj2TP+e1xFSfk=
golang.org/x/time v0.15.0/go.mod h1:Y4YMaQmXwGQZoFaVFk4YpCt4FLQMYKZe9oeV/f4MSno= golang.org/x/time v0.5.0/go.mod h1:3BpzKBy/shNhVucY/MWOyx10tF3SFh9QdLuxbVysPQM=
golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ= golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo= golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
golang.org/x/tools v0.0.0-20200130002326-2f3ba24bd6e7/go.mod h1:TB2adYChydJhpapKDTa4BR/hXlZSLoq2Wpct/0txZ28= golang.org/x/tools v0.0.0-20200130002326-2f3ba24bd6e7/go.mod h1:TB2adYChydJhpapKDTa4BR/hXlZSLoq2Wpct/0txZ28=
@@ -245,5 +243,5 @@ gopkg.in/yaml.v2 v2.3.0/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM= gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA= gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM= gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
gvisor.dev/gvisor v0.0.0-20261014193016-04d31b385980 h1:ribyrTYQ/QYeUIs8x8bVLYwGyepa7RnHLqJzegT5YYk= gvisor.dev/gvisor v0.0.0-20240423190808-9d7a357edefe h1:fre4i6mv4iBuz5lCMOzHD1rH1ljqHWSICFmZRbbgp3g=
gvisor.dev/gvisor v0.0.0-20261014193016-04d31b385980/go.mod h1:pK/dapF5Bf3KGW7/kd5CI5IC4lYh9FMFY1qETAIGLXk= gvisor.dev/gvisor v0.0.0-20240423190808-9d7a357edefe/go.mod h1:sxc3Uvk/vHcd3tj7/DHVBoR5wvWT/MmRq2pj7HRJnwU=
+5 -18
View File
@@ -31,7 +31,6 @@ package nebula
import ( import (
"bufio" "bufio"
"context"
"fmt" "fmt"
"net" "net"
"strconv" "strconv"
@@ -52,29 +51,17 @@ type graphiteConfigExport struct {
Percentiles []float64 // Percentiles to export from timers and histograms Percentiles []float64 // Percentiles to export from timers and histograms
} }
// graphiteTimeout bounds one submission, the connect and the write together. An export is a few KB, a host that can't
// take that in this long isn't going to.
const graphiteTimeout = 30 * time.Second
// graphiteOnce performs a single submission to Graphite, returning a non-nil // graphiteOnce performs a single submission to Graphite, returning a non-nil
// error if it could not connect or write all of it within timeout. It gives up // error on failed connections.
// as soon as ctx is done. func graphiteOnce(c graphiteConfigExport) error {
func graphiteOnce(ctx context.Context, c graphiteConfigExport, timeout time.Duration) error {
now := time.Now().Unix() now := time.Now().Unix()
du := float64(c.DurationUnit) du := float64(c.DurationUnit)
flushSeconds := float64(c.FlushInterval) / float64(time.Second) flushSeconds := float64(c.FlushInterval) / float64(time.Second)
deadline := time.Now().Add(timeout) conn, err := net.DialTCP("tcp", nil, c.Addr)
d := net.Dialer{Deadline: deadline}
conn, err := d.DialContext(ctx, "tcp", c.Addr.String())
if err != nil { if err != nil {
return err return err
} }
defer conn.Close() defer conn.Close()
if err := conn.SetDeadline(deadline); err != nil {
return err
}
stop := context.AfterFunc(ctx, func() { _ = conn.Close() })
defer stop()
w := bufio.NewWriter(conn) w := bufio.NewWriter(conn)
c.Registry.Each(func(name string, i any) { c.Registry.Each(func(name string, i any) {
switch metric := i.(type) { switch metric := i.(type) {
@@ -124,7 +111,7 @@ func graphiteOnce(ctx context.Context, c graphiteConfigExport, timeout time.Dura
fmt.Fprintf(w, "%s.%s.fifteen-minute %.2f %d\n", c.Prefix, name, t.Rate15(), now) fmt.Fprintf(w, "%s.%s.fifteen-minute %.2f %d\n", c.Prefix, name, t.Rate15(), now)
fmt.Fprintf(w, "%s.%s.mean-rate %.2f %d\n", c.Prefix, name, t.RateMean(), now) fmt.Fprintf(w, "%s.%s.mean-rate %.2f %d\n", c.Prefix, name, t.RateMean(), now)
} }
w.Flush()
}) })
// A failed write sticks, so this reports the first one return nil
return w.Flush()
} }
-315
View File
@@ -1,315 +0,0 @@
package nebula
import (
"bytes"
"context"
"errors"
"fmt"
"io"
"net"
"strings"
"sync"
"testing"
"time"
"github.com/rcrowley/go-metrics"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// graphiteHost accepts connections and hands each one to the test
type graphiteHost struct {
ln net.Listener
conns chan net.Conn
mu sync.Mutex
closed bool
open []net.Conn
}
// newGraphiteHost accepts connections for the test. A non-zero window shrinks each connection's receive buffer, so a
// sender that is not read from blocks after a little whatever the host's TCP tuning. Only for tests that need a stall,
// a small window makes reading very slow on Linux.
func newGraphiteHost(t *testing.T, window int) *graphiteHost {
t.Helper()
ln, err := net.Listen("tcp", "127.0.0.1:0")
require.NoError(t, err)
h := &graphiteHost{ln: ln, conns: make(chan net.Conn, 8)}
done := make(chan struct{})
t.Cleanup(func() {
_ = ln.Close()
h.mu.Lock()
h.closed = true
for _, c := range h.open {
_ = c.Close()
}
h.mu.Unlock()
close(done)
})
go func() {
for {
c, err := ln.Accept()
if err != nil {
return
}
h.mu.Lock()
if h.closed {
h.mu.Unlock()
_ = c.Close()
return
}
h.open = append(h.open, c)
h.mu.Unlock()
if tc, ok := c.(*net.TCPConn); ok && window > 0 {
_ = tc.SetReadBuffer(window)
}
select {
case h.conns <- c:
case <-done:
return
}
}
}()
return h
}
func (h *graphiteHost) addr(t *testing.T) *net.TCPAddr {
a, err := net.ResolveTCPAddr("tcp", h.ln.Addr().String())
require.NoError(t, err)
return a
}
// readExport reads what one send wrote, bounded so a send that never finishes fails the test rather than hanging it
func readExport(t *testing.T, c net.Conn) []byte {
t.Helper()
require.NoError(t, c.SetReadDeadline(time.Now().Add(10*time.Second)))
b, err := io.ReadAll(c)
require.NoError(t, err)
return b
}
func (h *graphiteHost) next(t *testing.T, wait time.Duration) net.Conn {
t.Helper()
select {
case c := <-h.conns:
return c
case <-time.After(wait):
t.Fatal("no connection")
return nil
}
}
// lockedBuffer is a log destination a test can read while the code under test writes to it from another goroutine
type lockedBuffer struct {
mu sync.Mutex
buf bytes.Buffer
}
func (b *lockedBuffer) Write(p []byte) (int, error) {
b.mu.Lock()
defer b.mu.Unlock()
return b.buf.Write(p)
}
func (b *lockedBuffer) String() string {
b.mu.Lock()
defer b.mu.Unlock()
return b.buf.String()
}
func testGraphiteConfig(addr *net.TCPAddr, r metrics.Registry) graphiteConfigExport {
return graphiteConfigExport{Addr: addr, Registry: r, FlushInterval: time.Second, DurationUnit: time.Nanosecond, Prefix: "nebula"}
}
// Enough metrics that one export overflows the socket buffers on any platform, so a send to a host that stops
// reading blocks
func bigGraphiteRegistry(r metrics.Registry, t *testing.T) {
for i := range 200000 {
name := fmt.Sprintf("big.gauge.%d", i)
metrics.GetOrRegisterGauge(name, r).Update(int64(i))
t.Cleanup(func() { r.Unregister(name) })
}
}
func TestGraphiteOnce_sends(t *testing.T) {
r := metrics.NewRegistry()
metrics.GetOrRegisterGauge("g", r).Update(7)
metrics.GetOrRegisterCounter("c", r).Inc(3)
h := newGraphiteHost(t, 0)
require.NoError(t, graphiteOnce(t.Context(), testGraphiteConfig(h.addr(t), r), time.Minute))
b := readExport(t, h.next(t, 5*time.Second))
assert.Contains(t, string(b), "nebula.g.value 7 ")
assert.Contains(t, string(b), "nebula.c.count 3 ")
assert.Contains(t, string(b), "nebula.c.count_ps 3.00 ")
}
// A host that accepts and never reads is given up on, and the failed write is reported
func TestGraphiteOnce_stalledHostTimesOut(t *testing.T) {
r := metrics.NewRegistry()
bigGraphiteRegistry(r, t)
h := newGraphiteHost(t, 4096)
done := make(chan error, 1)
cfg := testGraphiteConfig(h.addr(t), r)
go func() { done <- graphiteOnce(t.Context(), cfg, 300*time.Millisecond) }()
h.next(t, 5*time.Second) // accepted, never read
select {
case err := <-done:
require.Error(t, err)
case <-time.After(10 * time.Second):
t.Fatal("the send did not give up on a host that stopped reading")
}
}
// The timeout covers the whole send, a host that keeps reading but too slowly is still given up on
func TestGraphiteOnce_timeoutCoversTheWholeSend(t *testing.T) {
r := metrics.NewRegistry()
bigGraphiteRegistry(r, t)
h := newGraphiteHost(t, 0)
done := make(chan error, 1)
start := time.Now()
cfg := testGraphiteConfig(h.addr(t), r)
go func() { done <- graphiteOnce(t.Context(), cfg, 500*time.Millisecond) }()
c := h.next(t, 5*time.Second)
stopReading := make(chan struct{})
defer close(stopReading)
// 32KB every 10ms keeps every write well inside the timeout, the whole export takes seconds
go func() {
buf := make([]byte, 32<<10)
for {
select {
case <-stopReading:
return
case <-time.After(10 * time.Millisecond):
}
if _, err := c.Read(buf); err != nil {
return
}
}
}()
select {
case err := <-done:
require.Error(t, err)
assert.Less(t, time.Since(start), 5*time.Second)
case <-time.After(10 * time.Second):
t.Fatal("a host that kept reading slowly held the send past its timeout")
}
}
func TestGraphiteOnce_cancelledCtxDoesNotConnect(t *testing.T) {
h := newGraphiteHost(t, 0)
ctx, cancel := context.WithCancel(t.Context())
cancel()
require.Error(t, graphiteOnce(ctx, testGraphiteConfig(h.addr(t), metrics.NewRegistry()), time.Minute))
select {
case <-h.conns:
t.Fatal("connected on a cancelled ctx")
case <-time.After(200 * time.Millisecond):
}
}
// A stop or reload abandons a stuck send rather than waiting out its timeout
func TestGraphiteOnce_cancelAbandonsSend(t *testing.T) {
r := metrics.NewRegistry()
bigGraphiteRegistry(r, t)
h := newGraphiteHost(t, 4096)
ctx, cancel := context.WithCancel(t.Context())
done := make(chan error, 1)
cfg := testGraphiteConfig(h.addr(t), r)
go func() { done <- graphiteOnce(ctx, cfg, time.Minute) }()
h.next(t, 5*time.Second)
select {
case err := <-done:
t.Fatalf("the send should be stuck on a host that is not reading, it ended with %v", err)
case <-time.After(200 * time.Millisecond):
}
cancel()
select {
case err := <-done:
require.Error(t, err)
case <-time.After(5 * time.Second):
t.Fatal("the send is still stuck after its ctx was cancelled")
}
}
// buildRuntime hands graphite the configured prefix and interval
func TestStatsServer_graphiteExportUsesConfig(t *testing.T) {
const name = "graphite.wiring.test"
metrics.GetOrRegisterCounter(name, nil).Inc(8)
t.Cleanup(func() { metrics.DefaultRegistry.Unregister(name) })
h := newGraphiteHost(t, 0)
s, c := newTestStatsServer(t)
setStatsConfig(c, map[string]any{"type": "graphite", "interval": "4s", "prefix": "pfx", "host": h.ln.Addr().String()})
cfg, err := loadStatsConfig(c)
require.NoError(t, err)
s.runMu.Lock()
fns, _ := s.buildRuntime(t.Context(), cfg)
s.runMu.Unlock()
for _, fn := range fns {
fn()
}
b := readExport(t, h.next(t, 5*time.Second))
assert.Contains(t, string(b), "pfx."+name+".count 8 ")
assert.Contains(t, string(b), "pfx."+name+".count_ps 2.00 ")
}
func TestStatsServer_graphiteFailureIsLogged(t *testing.T) {
logs := &lockedBuffer{}
s, c := newTestStatsServer(t)
s.l = test.NewLoggerWithOutput(logs)
setStatsConfig(c, map[string]any{"type": "graphite", "interval": "1h", "host": "127.0.0.1:" + freeTCPPort(t)})
cfg, err := loadStatsConfig(c)
require.NoError(t, err)
s.runMu.Lock()
fns, _ := s.buildRuntime(t.Context(), cfg)
s.runMu.Unlock()
for _, fn := range fns {
fn()
}
assert.Contains(t, logs.String(), "Graphite export failed")
}
// Stop abandons the running export, so it goes through Start rather than handing buildRuntime a ctx, and a stop is
// not logged as a failure
func TestStatsServer_stopAbandonsGraphiteSend(t *testing.T) {
bigGraphiteRegistry(metrics.DefaultRegistry, t)
h := newGraphiteHost(t, 4096)
logs := &lockedBuffer{}
s, c := newTestStatsServer(t)
s.l = test.NewLoggerWithOutput(logs)
setStatsConfig(c, map[string]any{"type": "graphite", "interval": "50ms", "host": h.ln.Addr().String()})
require.NoError(t, s.reload(c, true))
started := make(chan struct{})
go func() {
s.Start()
close(started)
}()
stuck := h.next(t, 30*time.Second)
s.Stop()
select {
case <-started:
case <-time.After(5 * time.Second):
t.Fatal("Start did not return after Stop")
}
// A closed sender resets the connection once it sees this, one still sending just buffers it. Either way the
// read below doesn't have to drain the export through the small window to find out
_, err := stuck.Write([]byte("x"))
require.NoError(t, err)
require.NoError(t, stuck.SetReadDeadline(time.Now().Add(10*time.Second)))
n, err := io.Copy(io.Discard, stuck)
var ne net.Error
if errors.As(err, &ne) && ne.Timeout() {
t.Fatalf("the send kept the connection after Stop, read %d bytes", n)
}
assert.Less(t, n, int64(1<<20), "the send kept going after Stop")
// The capture loop logs on its own goroutine after Start has returned, give it the chance
assert.Never(t, func() bool { return strings.Contains(logs.String(), "Graphite export failed") }, 500*time.Millisecond, 10*time.Millisecond, "a stop is not a failure")
}
+35 -42
View File
@@ -54,24 +54,22 @@ type HandshakeConfig struct {
} }
type HandshakeManager struct { type HandshakeManager struct {
// Mutex for interacting with the vpnIps and indexes maps. // Mutex for interacting with the vpnIps and indexes maps
// Lock order is hostmap, then this lock, then the lighthouse, then a RemoteList.
sync.RWMutex sync.RWMutex
vpnIps map[netip.Addr]*HandshakeHostInfo vpnIps map[netip.Addr]*HandshakeHostInfo
indexes map[uint32]*HandshakeHostInfo indexes map[uint32]*HandshakeHostInfo
mainHostMap *HostMap mainHostMap *HostMap
lightHouse *LightHouse lightHouse *LightHouse
outside udp.Conn outside udp.Conn
config HandshakeConfig config HandshakeConfig
OutboundHandshakeTimer *LockingTimerWheel[netip.Addr] OutboundHandshakeTimer *LockingTimerWheel[netip.Addr]
messageMetrics *MessageMetrics messageMetrics *MessageMetrics
metricInitiated metrics.Counter metricInitiated metrics.Counter
metricTimedOut metrics.Counter metricTimedOut metrics.Counter
metricRejectedBlocklisted metrics.Counter f *Interface
f *Interface l *slog.Logger
l *slog.Logger
// can be used to trigger outbound handshake for the given vpnIp // can be used to trigger outbound handshake for the given vpnIp
trigger chan netip.Addr trigger chan netip.Addr
@@ -119,19 +117,18 @@ func (hh *HandshakeHostInfo) cachePacket(l *slog.Logger, t header.MessageType, s
func NewHandshakeManager(l *slog.Logger, mainHostMap *HostMap, lightHouse *LightHouse, outside udp.Conn, config HandshakeConfig) *HandshakeManager { func NewHandshakeManager(l *slog.Logger, mainHostMap *HostMap, lightHouse *LightHouse, outside udp.Conn, config HandshakeConfig) *HandshakeManager {
return &HandshakeManager{ return &HandshakeManager{
vpnIps: map[netip.Addr]*HandshakeHostInfo{}, vpnIps: map[netip.Addr]*HandshakeHostInfo{},
indexes: map[uint32]*HandshakeHostInfo{}, indexes: map[uint32]*HandshakeHostInfo{},
mainHostMap: mainHostMap, mainHostMap: mainHostMap,
lightHouse: lightHouse, lightHouse: lightHouse,
outside: outside, outside: outside,
config: config, config: config,
trigger: make(chan netip.Addr, config.triggerBuffer), trigger: make(chan netip.Addr, config.triggerBuffer),
OutboundHandshakeTimer: NewLockingTimerWheel[netip.Addr](config.tryInterval, hsTimeout(config.retries, config.tryInterval)), OutboundHandshakeTimer: NewLockingTimerWheel[netip.Addr](config.tryInterval, hsTimeout(config.retries, config.tryInterval)),
messageMetrics: config.messageMetrics, messageMetrics: config.messageMetrics,
metricInitiated: metrics.GetOrRegisterCounter("handshake_manager.initiated", nil), metricInitiated: metrics.GetOrRegisterCounter("handshake_manager.initiated", nil),
metricTimedOut: metrics.GetOrRegisterCounter("handshake_manager.timed_out", nil), metricTimedOut: metrics.GetOrRegisterCounter("handshake_manager.timed_out", nil),
metricRejectedBlocklisted: metrics.GetOrRegisterCounter("handshake_manager.rejected.blocklisted", nil), l: l,
l: l,
} }
} }
@@ -262,6 +259,14 @@ func (hm *HandshakeManager) handleOutbound(vpnIp netip.Addr, lighthouseTriggered
"style": header.SubTypeName(header.Handshake, hh.machine.Subtype()), "style": header.SubTypeName(header.Handshake, hh.machine.Subtype()),
} }
// Get a remotes object if we don't already have one.
// This is mainly to protect us as this should never be the case
// NB ^ This comment doesn't jive. It's how the thing gets initialized.
// It's the common path. Should it update every time, in case a future LH query/queries give us more info?
if hostinfo.remotes == nil {
hostinfo.remotes = hm.lightHouse.QueryCache([]netip.Addr{vpnIp})
}
remotes := hostinfo.remotes.CopyAddrs(hm.mainHostMap.GetPreferredRanges()) remotes := hostinfo.remotes.CopyAddrs(hm.mainHostMap.GetPreferredRanges())
remotesHaveChanged := !slices.Equal(remotes, hh.lastRemotes) remotesHaveChanged := !slices.Equal(remotes, hh.lastRemotes)
@@ -366,9 +371,6 @@ func (hm *HandshakeManager) StartHandshake(vpnAddr netip.Addr, cacheCb func(*Han
hostinfo := &HostInfo{ hostinfo := &HostInfo{
vpnAddrs: []netip.Addr{vpnAddr}, vpnAddrs: []netip.Addr{vpnAddr},
HandshakePacket: make(map[uint8][]byte, 0), HandshakePacket: make(map[uint8][]byte, 0),
// Attached here so a pending hostinfo never exists without one: SetRemote dereferences
// it, and create-tunnel -address can reach SetRemote before the handshake loop runs.
remotes: hm.lightHouse.QueryCache([]netip.Addr{vpnAddr}),
relayState: RelayState{ relayState: RelayState{
relays: nil, relays: nil,
relayForByAddr: map[netip.Addr]*Relay{}, relayForByAddr: map[netip.Addr]*Relay{},
@@ -719,12 +721,7 @@ func (hm *HandshakeManager) beginHandshake(via ViaSender, packet []byte, h *head
response, result, err := machine.ProcessPacket(nil, packet) response, result, err := machine.ProcessPacket(nil, packet)
if err != nil { if err != nil {
level := slog.LevelError f.l.Error("Failed to process handshake packet", "from", via, "error", err)
if errors.Is(err, cert.ErrBlockListed) {
level = slog.LevelDebug
hm.metricRejectedBlocklisted.Inc(1)
}
f.l.Log(context.Background(), level, "Failed to process handshake packet", "from", via, "error", err)
return return
} }
@@ -856,9 +853,6 @@ func (hm *HandshakeManager) continueHandshake(via ViaSender, hh *HandshakeHostIn
response, result, err := machine.ProcessPacket(nil, packet) response, result, err := machine.ProcessPacket(nil, packet)
if err != nil { if err != nil {
if errors.Is(err, cert.ErrBlockListed) {
hm.metricRejectedBlocklisted.Inc(1)
}
// Recoverable errors are routine noise, log at Debug. Fatal errors get a Warn. // Recoverable errors are routine noise, log at Debug. Fatal errors get a Warn.
if machine.Failed() { if machine.Failed() {
f.l.Warn("Failed to process handshake packet, abandoning", f.l.Warn("Failed to process handshake packet, abandoning",
@@ -952,14 +946,13 @@ func (hm *HandshakeManager) continueHandshake(via ViaSender, hh *HandshakeHostIn
hm.DeleteHostInfo(hostinfo) hm.DeleteHostInfo(hostinfo)
hm.StartHandshake(hostinfo.vpnAddrs[0], func(newHH *HandshakeHostInfo) { hm.StartHandshake(hostinfo.vpnAddrs[0], func(newHH *HandshakeHostInfo) {
newHH.hostinfo.remotes = hostinfo.remotes
newHH.hostinfo.remotes.BlockRemote(via) newHH.hostinfo.remotes.BlockRemote(via)
newHH.packetStore = hh.packetStore newHH.packetStore = hh.packetStore
hh.packetStore = []*cachedPacket{} hh.packetStore = []*cachedPacket{}
hostinfo.vpnAddrs = vpnAddrs
f.sendCloseTunnel(hostinfo)
}) })
// sendCloseTunnel may take the hostmap lock, so it must not run under the handshake
// manager lock that StartHandshake holds around the callback.
hostinfo.vpnAddrs = vpnAddrs
f.sendCloseTunnel(hostinfo)
return return
} }
+2 -156
View File
@@ -1,22 +1,16 @@
package nebula package nebula
import ( import (
"log/slog"
"net/netip" "net/netip"
"testing" "testing"
"time" "time"
"github.com/flynn/noise"
"github.com/gaissmai/bart" "github.com/gaissmai/bart"
"github.com/rcrowley/go-metrics"
"github.com/slackhq/nebula/cert" "github.com/slackhq/nebula/cert"
ct "github.com/slackhq/nebula/cert_test"
"github.com/slackhq/nebula/handshake"
"github.com/slackhq/nebula/header" "github.com/slackhq/nebula/header"
"github.com/slackhq/nebula/test" "github.com/slackhq/nebula/test"
"github.com/slackhq/nebula/udp" "github.com/slackhq/nebula/udp"
"github.com/stretchr/testify/assert" "github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
) )
func Test_NewHandshakeManagerVpnIp(t *testing.T) { func Test_NewHandshakeManagerVpnIp(t *testing.T) {
@@ -48,6 +42,8 @@ func Test_NewHandshakeManagerVpnIp(t *testing.T) {
i2 := blah.StartHandshake(ip, nil) i2 := blah.StartHandshake(ip, nil)
assert.Same(t, i, i2) assert.Same(t, i, i2)
i.remotes = NewRemoteList([]netip.Addr{}, nil)
// Adding something to pending should not affect the main hostmap // Adding something to pending should not affect the main hostmap
assert.Empty(t, mainHM.Hosts) assert.Empty(t, mainHM.Hosts)
@@ -239,153 +235,3 @@ func TestHandleIncomingDispatch(t *testing.T) {
assert.Empty(t, hm.indexes, "orphan stage-2 must not create state") assert.Empty(t, hm.indexes, "orphan stage-2 must not create state")
}) })
} }
func TestBlocklistedHandshakeLogLevelAndCounter(t *testing.T) {
ca, _, caKey, _ := ct.NewTestCaCert(cert.Version2, cert.Curve_CURVE25519, time.Time{}, time.Time{}, nil, nil, nil)
pool := ct.NewTestCAPool(ca)
newCredential := func(name, network string) (cert.Certificate, *handshake.Credential) {
c, _, rawKey, _ := ct.NewTestCert(cert.Version2, cert.Curve_CURVE25519, ca, caKey,
name, ca.NotBefore(), ca.NotAfter(), []netip.Prefix{netip.MustParsePrefix(network)}, nil, nil)
privateKey, _, _, err := cert.UnmarshalPrivateKeyFromPEM(rawKey)
require.NoError(t, err)
certBytes, err := c.MarshalForHandshakes()
require.NoError(t, err)
cipherSuite := noise.NewCipherSuite(noise.DH25519, noise.CipherChaChaPoly, noise.HashSHA256)
return c, handshake.NewCredential(c, certBytes, privateKey, cipherSuite)
}
initiatorCert, initiatorCredential := newCredential("initiator", "10.0.0.1/24")
responderCert, responderCredential := newCredential("responder", "10.0.0.2/24")
fingerprint, err := initiatorCert.Fingerprint()
require.NoError(t, err)
pool.BlocklistFingerprint(fingerprint)
initiator, err := handshake.NewMachine(cert.Version2,
func(cert.Version) *handshake.Credential { return initiatorCredential },
func(c cert.Certificate) (*cert.CachedCertificate, error) {
return pool.VerifyCertificate(time.Now(), c)
},
func() (uint32, error) { return 1, nil }, true, header.HandshakeIXPSK0)
require.NoError(t, err)
packet, err := initiator.Initiate(nil)
require.NoError(t, err)
handler := &capturingHandler{}
logger := slog.New(handler)
pki := &PKI{}
pki.cs.Store(&CertState{initiatingVersion: cert.Version2, v2Credential: responderCredential})
pki.caPool.Store(pool)
hm := NewHandshakeManager(logger, nil, nil, nil, defaultHandshakeConfig)
hm.metricRejectedBlocklisted = metrics.NewCounter()
hm.f = &Interface{pki: pki, l: logger}
via := ViaSender{UdpAddr: netip.MustParseAddrPort("198.51.100.7:4242")}
hm.beginHandshake(via, packet, nil)
entry := findEntry(t, handler.entries, "Failed to process handshake packet")
assert.Equal(t, slog.LevelDebug, entry.Level)
require.ErrorIs(t, entry.Attrs["error"].(error), cert.ErrBlockListed)
assert.Equal(t, int64(1), hm.metricRejectedBlocklisted.Count())
handler.entries = nil
hm.beginHandshake(via, nil, nil)
entry = findEntry(t, handler.entries, "Failed to process handshake packet")
assert.Equal(t, slog.LevelError, entry.Level)
assert.Equal(t, int64(1), hm.metricRejectedBlocklisted.Count())
// The initiator sees the same rejection in continueHandshake, after a
// responder has accepted its first packet and sent a valid reply.
cleanPool := ct.NewTestCAPool(ca)
responder, err := handshake.NewMachine(cert.Version2,
func(cert.Version) *handshake.Credential { return responderCredential },
func(c cert.Certificate) (*cert.CachedCertificate, error) {
return cleanPool.VerifyCertificate(time.Now(), c)
},
func() (uint32, error) { return 2, nil }, false, header.HandshakeIXPSK0)
require.NoError(t, err)
response, _, err := responder.ProcessPacket(nil, packet)
require.NoError(t, err)
responderFingerprint, err := responderCert.Fingerprint()
require.NoError(t, err)
pool.BlocklistFingerprint(responderFingerprint)
initiatingHM := NewHandshakeManager(logger, nil, nil, nil, defaultHandshakeConfig)
initiatingHM.metricRejectedBlocklisted = metrics.NewCounter()
initiatingHM.f = &Interface{l: logger}
hh := &HandshakeHostInfo{
machine: initiator,
hostinfo: &HostInfo{localIndexId: 1, vpnAddrs: []netip.Addr{netip.MustParseAddr("10.0.0.2")}},
}
initiatingHM.indexes[1] = hh
via.IsRelayed = true // skip the remote allow list before certificate validation
initiatingHM.continueHandshake(via, hh, response)
assert.Equal(t, int64(1), initiatingHM.metricRejectedBlocklisted.Count())
}
// Test_PendingHandshakeHasRemotes covers create-tunnel -address: a pending hostinfo must already
// carry its RemoteList when StartHandshake runs its callback, so SetRemote there, before the
// handshake loop has run, is safe and the address becomes one the handshake is sent to.
func Test_PendingHandshakeHasRemotes(t *testing.T) {
l := test.NewLogger()
mainHM := newHostMap(l)
preferredRanges := []netip.Prefix{}
mainHM.preferredRanges.Store(&preferredRanges)
hm := NewHandshakeManager(l, mainHM, newTestLighthouse(), &udp.NoopConn{}, defaultHandshakeConfig)
vpnIp := netip.MustParseAddr("172.1.1.2")
remote := netip.MustParseAddrPort("192.168.1.9:4242")
hi := hm.StartHandshake(vpnIp, func(hh *HandshakeHostInfo) {
require.NotNil(t, hh.hostinfo.remotes, "a pending handshake must never be without its RemoteList")
hh.hostinfo.SetRemote(remote)
})
assert.Contains(t, hi.remotes.CopyAddrs(preferredRanges), remote, "the handshake must be sent to the supplied remote")
assert.Equal(t, remote, hi.GetRemote())
// A RemoteList keeps one learned address per host, so a second remote replaces the
// first — the same behaviour change-remote has on an established tunnel.
other := netip.MustParseAddrPort("192.168.1.10:4242")
hi.SetRemote(other)
assert.Equal(t, []netip.AddrPort{other}, hi.remotes.CopyAddrs(preferredRanges))
assert.Equal(t, other, hi.GetRemote())
}
// newRemotesRaceFixture builds a handshake manager with an established tunnel to vpnAddr whose
// RemoteList is the lighthouse's cached list, the state every teardown-during-handshake race starts from.
func newRemotesRaceFixture(t *testing.T, vpnAddr netip.Addr) (*HandshakeManager, *Interface, *HostInfo) {
t.Helper()
l := test.NewLogger()
mainHM := newHostMap(l)
preferredRanges := []netip.Prefix{}
mainHM.preferredRanges.Store(&preferredRanges)
lh := newTestLighthouse()
hm := NewHandshakeManager(l, mainHM, lh, &udp.NoopConn{}, defaultHandshakeConfig)
f := &Interface{hostMap: mainHM, lightHouse: lh, handshakeManager: hm, l: l}
hm.f = f
established := &HostInfo{
vpnAddrs: []netip.Addr{vpnAddr},
localIndexId: 1,
remoteIndexId: 101,
remotes: lh.QueryCache([]netip.Addr{vpnAddr}),
HandshakePacket: map[uint8][]byte{},
}
mainHM.Lock()
mainHM.unlockedAddHostInfo(established, f)
mainHM.Unlock()
return hm, f, established
}
// With no handshake pending, tearing down the last tunnel still evicts the list.
func Test_CloseTunnelEvictsWithoutPendingHandshake(t *testing.T) {
vpnAddr := netip.MustParseAddr("172.1.1.2")
_, f, established := newRemotesRaceFixture(t, vpnAddr)
f.closeTunnel(established)
assert.NotContains(t, f.hostMap.Hosts, vpnAddr)
assert.NotContains(t, f.lightHouse.addrMap, vpnAddr)
}
-5
View File
@@ -24,11 +24,6 @@ type m = map[string]any
const ( const (
Version uint8 = 1 Version uint8 = 1
Len = 16 Len = 16
// MaxOverhead is the most a payload grows on its way to the underlay.
// A relayed packet has two headers and two AEAD tags, one for the inner tunnel and another for the relay tunnel.
// Every supported cipher (today) has a 16 byte tag.
MaxOverhead = 2 * (Len + 16)
) )
type MessageType uint8 type MessageType uint8
+5 -7
View File
@@ -316,9 +316,9 @@ type cachedPacketMetrics struct {
func NewHostMapFromConfig(l *slog.Logger, c *config.C) *HostMap { func NewHostMapFromConfig(l *slog.Logger, c *config.C) *HostMap {
hm := newHostMap(l) hm := newHostMap(l)
hm.reload(c, true, nil) hm.reload(c, true)
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) { c.RegisterReloadCallback(func(c *config.C) {
hm.reload(c, false, r) hm.reload(c, false)
}) })
l.Info("Main HostMap created", "preferredRanges", hm.GetPreferredRanges()) l.Info("Main HostMap created", "preferredRanges", hm.GetPreferredRanges())
@@ -337,8 +337,7 @@ func newHostMap(l *slog.Logger) *HostMap {
} }
} }
// r is nil on the initial load func (hm *HostMap) reload(c *config.C, initial bool) {
func (hm *HostMap) reload(c *config.C, initial bool, r *config.Reload) {
if initial || c.HasChanged("preferred_ranges") { if initial || c.HasChanged("preferred_ranges") {
var preferredRanges []netip.Prefix var preferredRanges []netip.Prefix
rawPreferredRanges := c.GetStringSlice("preferred_ranges", []string{}) rawPreferredRanges := c.GetStringSlice("preferred_ranges", []string{})
@@ -349,9 +348,8 @@ func (hm *HostMap) reload(c *config.C, initial bool, r *config.Reload) {
if err != nil { if err != nil {
hm.l.Warn("Failed to parse preferred ranges, ignoring", hm.l.Warn("Failed to parse preferred ranges, ignoring",
"error", err, "error", err,
"range", rawPreferredRange, "range", rawPreferredRanges,
) )
r.Failed()
continue continue
} }
-13
View File
@@ -447,16 +447,3 @@ func TestHostInfo_markOutOnly(t *testing.T) {
assert.True(t, h.markOut(6), "a relayed send must not consume the epoch edge") assert.True(t, h.markOut(6), "a relayed send must not consume the epoch edge")
assert.False(t, h.markOut(6)) assert.False(t, h.markOut(6))
} }
// A preferred range that doesn't parse is skipped and the rest are used, the reload says one was skipped
func TestHostMap_reloadBadPreferredRangeFails(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
require.NoError(t, c.LoadString("preferred_ranges: [10.0.0.0/8]\n"))
hm := NewHostMapFromConfig(l, c)
res := c.ReloadConfigString("preferred_ranges: [10.0.0.0/8, nope]\n")
require.NoError(t, res.Err)
assert.True(t, res.Failed)
assert.Equal(t, []netip.Prefix{netip.MustParsePrefix("10.0.0.0/8")}, hm.GetPreferredRanges())
}
+18 -35
View File
@@ -10,7 +10,6 @@ import (
"github.com/slackhq/nebula/firewall" "github.com/slackhq/nebula/firewall"
"github.com/slackhq/nebula/header" "github.com/slackhq/nebula/header"
"github.com/slackhq/nebula/iputil" "github.com/slackhq/nebula/iputil"
"github.com/slackhq/nebula/logging"
"github.com/slackhq/nebula/noiseutil" "github.com/slackhq/nebula/noiseutil"
"github.com/slackhq/nebula/overlay/batch" "github.com/slackhq/nebula/overlay/batch"
"github.com/slackhq/nebula/overlay/tio" "github.com/slackhq/nebula/overlay/tio"
@@ -199,8 +198,8 @@ func (f *Interface) sendInsideMessage(hostinfo *HostInfo, pkt tio.Packet, nb []b
} }
err = tio.SegmentSuperpacket(pkt, func(seg []byte) error { err = tio.SegmentSuperpacket(pkt, func(seg []byte) error {
//relay header + header + plaintext + AEAD tag + relay tag //relay header + header + plaintext + AEAD tag (16 bytes for both AES-GCM and ChaCha20-Poly1305) + relay tag
scratch := sendBatch.Reserve(len(seg) + header.MaxOverhead) scratch := sendBatch.Reserve(header.Len + header.Len + len(seg) + 16 + 16)
innerPacket := f.sendInsideEncrypt(hostinfo, ci, seg, scratch[header.Len:], nb) innerPacket := f.sendInsideEncrypt(hostinfo, ci, seg, scratch[header.Len:], nb)
if innerPacket == nil { if innerPacket == nil {
@@ -518,9 +517,7 @@ func (f *Interface) SendVia(via *HostInfo, relay *Relay, ad, nb, out []byte, noc
err = f.writers[q].WriteTo(toSend, via.GetRemote()) err = f.writers[q].WriteTo(toSend, via.GetRemote())
if err != nil { if err != nil {
if l, ok := logging.RateLimited(f.l, f.udpRelayTxErrLimiter); ok { via.logger(f.l).Info("Failed to WriteTo in sendVia", "error", err)
via.logger(l).Info("Failed to WriteTo in sendVia", "error", err)
}
} }
} }
@@ -529,26 +526,16 @@ func (f *Interface) sendNoMetrics(t header.MessageType, st header.MessageSubType
return return
} }
useRelay := !remote.IsValid() && !hostinfo.GetRemote().IsValid() useRelay := !remote.IsValid() && !hostinfo.GetRemote().IsValid()
// The encrypt appends to out in place and a relayed packet is then wrapped in place,
// so out has to fit the whole packet up front or the encrypt reallocates out from under SendVia.
// We cap the MTU of tun devices and routes we create ourselves, so only a host that set a bigger MTU than we allow gets here.
need := header.Len + len(p) + ci.eKey.Overhead()
if useRelay {
// The relay tunnel runs the same cipher, and adds its own header and tag
need += header.Len + ci.eKey.Overhead()
}
if cap(out) < need {
if l, ok := logging.RateLimited(f.l, f.txOversizeLimiter); ok {
hostinfo.logger(l).Error("Dropping outbound packet, out buffer not large enough", "outCap", cap(out), "payloadLen", len(p), "relayed", useRelay)
}
return
}
fullOut := out fullOut := out
if useRelay { if useRelay {
if len(out) < header.Len {
// out always has a capacity of mtu, but not always a length greater than the header.Len.
// Grow it to make sure the next operation works.
out = out[:header.Len]
}
// Save a header's worth of data at the front of the 'out' buffer. // Save a header's worth of data at the front of the 'out' buffer.
out = out[header.Len:header.Len] out = out[header.Len:]
} }
if noiseutil.EncryptLockNeeded { if noiseutil.EncryptLockNeeded {
@@ -595,22 +582,18 @@ func (f *Interface) sendNoMetrics(t header.MessageType, st header.MessageSubType
if remote.IsValid() { if remote.IsValid() {
err = f.writers[q].WriteTo(out, remote) err = f.writers[q].WriteTo(out, remote)
if err != nil { if err != nil {
if l, ok := logging.RateLimited(f.l, f.udpTxErrLimiter); ok { hostinfo.logger(f.l).Error("Failed to write outgoing packet",
hostinfo.logger(l).Error("Failed to write outgoing packet", "error", err,
"error", err, "udpAddr", remote,
"udpAddr", remote, )
)
}
} }
} else if hr := hostinfo.GetRemote(); hr.IsValid() { } else if hr := hostinfo.GetRemote(); hr.IsValid() {
err = f.writers[q].WriteTo(out, hr) err = f.writers[q].WriteTo(out, hr)
if err != nil { if err != nil {
if l, ok := logging.RateLimited(f.l, f.udpTxErrLimiter); ok { hostinfo.logger(f.l).Error("Failed to write outgoing packet",
hostinfo.logger(l).Error("Failed to write outgoing packet", "error", err,
"error", err, "udpAddr", hr,
"udpAddr", hr, )
)
}
} }
} else { } else {
// Try to send via a relay // Try to send via a relay
+37 -95
View File
@@ -24,7 +24,6 @@ import (
"github.com/slackhq/nebula/overlay" "github.com/slackhq/nebula/overlay"
"github.com/slackhq/nebula/overlay/batch" "github.com/slackhq/nebula/overlay/batch"
"github.com/slackhq/nebula/overlay/tio" "github.com/slackhq/nebula/overlay/tio"
"github.com/slackhq/nebula/ratelimit"
"github.com/slackhq/nebula/udp" "github.com/slackhq/nebula/udp"
) )
@@ -124,12 +123,6 @@ type Interface struct {
// reused (every udp.Conn ListenOut guarantees that ordering). // reused (every udp.Conn ListenOut guarantees that ordering).
batchers []*batch.MultiCoalescer batchers []*batch.MultiCoalescer
wg sync.WaitGroup wg sync.WaitGroup
// readers covers just the listenIn and listenOut goroutines, wg also holds the construction token Close releases
readers sync.WaitGroup
udpTxErrLimiter *ratelimit.Limiter // udp direct transmit error log limiter
udpRelayTxErrLimiter *ratelimit.Limiter // udp relay transmit error log limiter
txOversizeLimiter *ratelimit.Limiter // oversized outbound packet drop log limiter
// fatalErr holds the first unexpected reader error that caused shutdown. // fatalErr holds the first unexpected reader error that caused shutdown.
// nil means "no fatal error" (yet) // nil means "no fatal error" (yet)
@@ -205,19 +198,6 @@ func NewInterface(ctx context.Context, c *InterfaceConfig) (*Interface, error) {
return nil, errors.New("no connection manager") return nil, errors.New("no connection manager")
} }
udpTxErrLimiter, err := ratelimit.New(1, time.Second)
if err != nil {
return nil, err
}
udpRelayTxErrLimiter, err := ratelimit.New(1, time.Second)
if err != nil {
return nil, err
}
txOversizeLimiter, err := ratelimit.New(1, time.Second)
if err != nil {
return nil, err
}
if c.routines <= 1 { if c.routines <= 1 {
c.PinThreads = false //pinning is not useful unless there's more than one tun reader c.PinThreads = false //pinning is not useful unless there's more than one tun reader
} }
@@ -250,9 +230,6 @@ func NewInterface(ctx context.Context, c *InterfaceConfig) (*Interface, error) {
conntrackCacheTimeout: c.ConntrackCacheTimeout, conntrackCacheTimeout: c.ConntrackCacheTimeout,
cpuAffinity: c.CpuAffinity, cpuAffinity: c.CpuAffinity,
pinThreads: c.PinThreads, pinThreads: c.PinThreads,
udpTxErrLimiter: udpTxErrLimiter,
udpRelayTxErrLimiter: udpRelayTxErrLimiter,
txOversizeLimiter: txOversizeLimiter,
metricHandshakes: metrics.GetOrRegisterHistogram("handshakes", nil, metrics.NewExpDecaySample(1028, 0.015)), metricHandshakes: metrics.GetOrRegisterHistogram("handshakes", nil, metrics.NewExpDecaySample(1028, 0.015)),
metricTxDropped: metrics.GetOrRegisterCounter("udp.tx.dropped", nil), metricTxDropped: metrics.GetOrRegisterCounter("udp.tx.dropped", nil),
@@ -343,18 +320,14 @@ func (f *Interface) activate() error {
func (f *Interface) run() { func (f *Interface) run() {
// Launch n queues to read packets from udp // Launch n queues to read packets from udp
for i := 0; i < f.routines; i++ { for i := 0; i < f.routines; i++ {
f.readers.Add(1)
f.wg.Go(func() { f.wg.Go(func() {
defer f.readers.Done()
f.listenOut(i) f.listenOut(i)
}) })
} }
// Launch n queues to read packets from tun dev // Launch n queues to read packets from tun dev
for i := 0; i < f.routines; i++ { for i := 0; i < f.routines; i++ {
f.readers.Add(1)
f.wg.Go(func() { f.wg.Go(func() {
defer f.readers.Done()
f.listenIn(f.queues[i], i) f.listenIn(f.queues[i], i)
}) })
} }
@@ -515,11 +488,10 @@ func (f *Interface) flushSendBatch(sb *batch.SendBatch, q int) {
func (f *Interface) RegisterConfigChangeCallbacks(c *config.C) { func (f *Interface) RegisterConfigChangeCallbacks(c *config.C) {
c.RegisterReloadCallback(f.reloadFirewall) c.RegisterReloadCallback(f.reloadFirewall)
for _, reload := range []func(*config.C){ c.RegisterReloadCallback(f.reloadSendRecvError)
f.reloadSendRecvError, f.reloadAcceptRecvError, f.reloadDisconnectInvalid, f.reloadMisc, c.RegisterReloadCallback(f.reloadAcceptRecvError)
} { c.RegisterReloadCallback(f.reloadDisconnectInvalid)
c.RegisterReloadCallback(func(c *config.C, _ *config.Reload) { reload(c) }) c.RegisterReloadCallback(f.reloadMisc)
}
for _, udpConn := range f.writers { for _, udpConn := range f.writers {
c.RegisterReloadCallback(udpConn.ReloadConfig) c.RegisterReloadCallback(udpConn.ReloadConfig)
@@ -536,19 +508,18 @@ func (f *Interface) reloadDisconnectInvalid(c *config.C) {
} }
} }
func (f *Interface) reloadFirewall(c *config.C, r *config.Reload) { func (f *Interface) reloadFirewall(c *config.C) {
cs := f.pki.getCertState() cs := f.pki.getCertState()
curCert := cs.getCertificate(cert.Version2) curCert := cs.getCertificate(cert.Version2)
if curCert == nil { if curCert == nil {
curCert = cs.getCertificate(cert.Version1) curCert = cs.getCertificate(cert.Version1)
} }
// Compare with what the running firewall was built from, not the last config, so a firewall that failed to build // The firewall builds its routableNetworks set from the certificate's UnsafeNetworks at construction.
// is tried again. The cert is the one this reload brought, PKI registers its reload callback before this one // Check to see if that set has changed, and if so, rebuild the firewall.
src := c.Render("firewall") certUnsafeChanged := curCert != nil && !slices.Equal(curCert.UnsafeNetworks(), f.firewall.unsafeNetworks)
if src != "" && src == f.firewall.source && curCert != nil &&
slices.Equal(curCert.Networks(), f.firewall.assignedNetworks) && if !c.HasChanged("firewall") && !certUnsafeChanged {
slices.Equal(curCert.UnsafeNetworks(), f.firewall.unsafeNetworks) {
f.l.Debug("No firewall config change detected") f.l.Debug("No firewall config change detected")
return return
} }
@@ -556,7 +527,6 @@ func (f *Interface) reloadFirewall(c *config.C, r *config.Reload) {
fw, err := NewFirewallFromConfig(f.l, cs, c) fw, err := NewFirewallFromConfig(f.l, cs, c)
if err != nil { if err != nil {
f.l.Error("Error while creating firewall during reload", "error", err) f.l.Error("Error while creating firewall during reload", "error", err)
r.Failed()
return return
} }
@@ -654,15 +624,17 @@ func (f *Interface) reloadMisc(c *config.C) {
} }
} }
// statsEmitter updates the interface's gauges, the stats server runs it on its capture loop. func (f *Interface) emitStats(ctx context.Context, i time.Duration) {
func (f *Interface) statsEmitter() func() { ticker := time.NewTicker(i)
defer ticker.Stop()
udpStats := udp.NewUDPStatsEmitter(f.writers) udpStats := udp.NewUDPStatsEmitter(f.writers)
certExpirationGauge := metrics.GetOrRegisterGauge("certificate.ttl_seconds", nil) certExpirationGauge := metrics.GetOrRegisterGauge("certificate.ttl_seconds", nil)
certInitiatingVersion := metrics.GetOrRegisterGauge("certificate.initiating_version", nil) certInitiatingVersion := metrics.GetOrRegisterGauge("certificate.initiating_version", nil)
certMaxVersion := metrics.GetOrRegisterGauge("certificate.max_version", nil) certMaxVersion := metrics.GetOrRegisterGauge("certificate.max_version", nil)
return func() { emit := func() {
f.firewall.EmitStats() f.firewall.EmitStats()
f.handshakeManager.EmitStats() f.handshakeManager.EmitStats()
udpStats() udpStats()
@@ -679,6 +651,19 @@ func (f *Interface) statsEmitter() func() {
certMaxVersion.Update(int64(certState.v1Cert.Version())) certMaxVersion.Update(int64(certState.v1Cert.Version()))
} }
} }
// Prime gauges so a Prometheus scrape that lands before the first tick
// sees real values instead of the zero defaults (issue #907).
emit()
for {
select {
case <-ctx.Done():
return
case <-ticker.C:
emit()
}
}
} }
func (f *Interface) GetHostInfo(vpnIp netip.Addr) *HostInfo { func (f *Interface) GetHostInfo(vpnIp netip.Addr) *HostInfo {
@@ -689,14 +674,6 @@ func (f *Interface) GetCertState() *CertState {
return f.pki.getCertState() return f.pki.getCertState()
} }
// shutdowner is a udp.Conn or overlay.Device whose readers can be woken without releasing their fds
type shutdowner interface {
Shutdown() error
}
// readerShutdownTimeout bounds how long Close waits on woken readers before releasing their fds anyway
const readerShutdownTimeout = 5 * time.Second
// Close releases the interface's resources: the udp sockets and the tun device. // Close releases the interface's resources: the udp sockets and the tun device.
// It is idempotent and safe to call at any point in the lifecycle, including on an interface that never activated, // It is idempotent and safe to call at any point in the lifecycle, including on an interface that never activated,
// calls after the first return nil without doing anything. // calls after the first return nil without doing anything.
@@ -706,55 +683,20 @@ func (f *Interface) Close() error {
} }
var errs []error var errs []error
closeUdp := func(i int, u udp.Conn) {
if err := u.Close(); err != nil { // Release the udp readers
for i, u := range f.writers {
err := u.Close()
if err != nil {
f.l.Error("Error while closing udp socket", "error", err, "writer", i) f.l.Error("Error while closing udp socket", "error", err, "writer", i)
errs = append(errs, err) errs = append(errs, err)
} }
} }
closeTun := func() {
if err := f.inside.Close(); err != nil {
errs = append(errs, err)
}
}
// Wake every reader before releasing any fd, one that loops back into a syscall on a reused fd number never wakes. // Release the tun device (closing the tun also closes all readers)
// Anything without Shutdown is closed right away, as before. closeErr := f.inside.Close()
var deferred []func() if closeErr != nil {
for i, u := range f.writers { errs = append(errs, closeErr)
if s, ok := u.(shutdowner); ok {
if err := s.Shutdown(); err != nil {
f.l.Error("Error while shutting down udp socket", "error", err, "writer", i)
errs = append(errs, err)
}
deferred = append(deferred, func() { closeUdp(i, u) })
} else {
closeUdp(i, u)
}
}
if s, ok := f.inside.(shutdowner); ok {
if err := s.Shutdown(); err != nil {
errs = append(errs, err)
}
deferred = append(deferred, closeTun)
} else {
closeTun()
}
if len(deferred) > 0 {
readersDone := make(chan struct{})
go func() {
f.readers.Wait()
close(readersDone)
}()
select {
case <-readersDone:
case <-time.After(readerShutdownTimeout):
f.l.Warn("Readers did not stop, closing their fds anyway", "timeout", readerShutdownTimeout)
}
for _, closeFn := range deferred {
closeFn()
}
} }
// Release the construction token so waiters know the resources are gone // Release the construction token so waiters know the resources are gone
+14 -5
View File
@@ -3,6 +3,7 @@
package nebula package nebula
import ( import (
"context"
"net/netip" "net/netip"
"testing" "testing"
"time" "time"
@@ -17,8 +18,11 @@ import (
"github.com/stretchr/testify/require" "github.com/stretchr/testify/require"
) )
// Priming before the first tick (issue #907) is Start's job, see TestStatsServer_Start_primes. // Test_emitStats_primesGauges covers issue #907: a Prometheus scrape that
func Test_statsEmitter_setsGauges(t *testing.T) { // landed before the first ticker fire used to read 0 for the cert gauges.
// emitStats now primes the gauges before entering the ticker loop. We assert
// the gauge is zero before the first call and non-zero after.
func Test_emitStats_primesGauges(t *testing.T) {
defer metrics.DefaultRegistry.UnregisterAll() defer metrics.DefaultRegistry.UnregisterAll()
l := test.NewLogger() l := test.NewLogger()
@@ -52,12 +56,17 @@ func Test_statsEmitter_setsGauges(t *testing.T) {
ifce.pki.cs.Store(cs) ifce.pki.cs.Store(cs)
ttlGauge := metrics.GetOrRegisterGauge("certificate.ttl_seconds", nil) ttlGauge := metrics.GetOrRegisterGauge("certificate.ttl_seconds", nil)
require.Zero(t, ttlGauge.Value(), "gauge should be zero before the emitter runs") require.Zero(t, ttlGauge.Value(), "gauge should be zero before emitStats runs")
ifce.statsEmitter()() // Pre-cancel the context so emitStats returns after priming the gauges
// without ever reading from ticker.C. The one hour interval is just a
// belt-and-suspenders, the test does not expect the ticker to fire.
ctx, cancel := context.WithCancel(context.Background())
cancel()
ifce.emitStats(ctx, time.Hour)
ttl := ttlGauge.Value() ttl := ttlGauge.Value()
assert.Positive(t, ttl, "ttl gauge should be set by one run of the emitter") assert.Positive(t, ttl, "ttl gauge should be primed by emitStats before its first tick")
assert.LessOrEqual(t, ttl, int64(3600)) assert.LessOrEqual(t, ttl, int64(3600))
assert.Equal(t, int64(cert.Version1), metrics.GetOrRegisterGauge("certificate.initiating_version", nil).Value()) assert.Equal(t, int64(cert.Version1), metrics.GetOrRegisterGauge("certificate.initiating_version", nil).Value())
assert.Equal(t, int64(cert.Version1), metrics.GetOrRegisterGauge("certificate.max_version", nil).Value()) assert.Equal(t, int64(cert.Version1), metrics.GetOrRegisterGauge("certificate.max_version", nil).Value())
+4 -81
View File
@@ -65,10 +65,10 @@ func TestReloadFirewall_CertUnsafeNetworksChanged(t *testing.T) {
pki.cs.Store(&CertState{v2Cert: c2, initiatingVersion: cert.Version2}) pki.cs.Store(&CertState{v2Cert: c2, initiatingVersion: cert.Version2})
// Reload with the same YAML so HasChanged("firewall") reports false. // Reload with the same YAML so HasChanged("firewall") reports false.
require.NoError(t, cfg.ReloadConfigString(rawYAML).Err) require.NoError(t, cfg.ReloadConfigString(rawYAML))
require.False(t, cfg.HasChanged("firewall")) require.False(t, cfg.HasChanged("firewall"))
f.reloadFirewall(cfg, nil) f.reloadFirewall(cfg)
assert.NotSame(t, fw, f.firewall, "firewall pointer should have been replaced") assert.NotSame(t, fw, f.firewall, "firewall pointer should have been replaced")
assert.Equal(t, newUnsafe, f.firewall.unsafeNetworks) assert.Equal(t, newUnsafe, f.firewall.unsafeNetworks)
@@ -113,85 +113,8 @@ func TestReloadFirewall_NoChange(t *testing.T) {
l: l, l: l,
} }
require.NoError(t, cfg.ReloadConfigString(rawYAML).Err) require.NoError(t, cfg.ReloadConfigString(rawYAML))
f.reloadFirewall(cfg, nil) f.reloadFirewall(cfg)
assert.Same(t, fw, f.firewall, "firewall should not have been replaced") assert.Same(t, fw, f.firewall, "firewall should not have been replaced")
} }
// Rules that don't parse keep the old firewall, and the reload says it could not apply them
func TestReloadFirewall_BadRulesFailTheReload(t *testing.T) {
l := test.NewLogger()
pki := &PKI{}
pki.cs.Store(&CertState{v2Cert: &dummyCert{version: cert.Version2, networks: []netip.Prefix{netip.MustParsePrefix("10.0.0.1/24")}}, initiatingVersion: cert.Version2})
rules := func(proto string) string {
return "firewall:\n outbound:\n - port: any\n proto: any\n host: any\n inbound:\n - port: any\n proto: " + proto + "\n host: any\n"
}
cfg := config.NewC(l)
require.NoError(t, cfg.LoadString(rules("any")))
fw, err := NewFirewallFromConfig(l, pki.getCertState(), cfg)
require.NoError(t, err)
f := &Interface{pki: pki, firewall: fw, l: l}
cfg.RegisterReloadCallback(f.reloadFirewall)
// The running firewall still isn't built from what the config asks for, so the same reload fails again
for range 2 {
res := cfg.ReloadConfigString(rules("nope"))
require.NoError(t, res.Err)
assert.True(t, res.Failed)
assert.Same(t, fw, f.firewall)
}
res := cfg.ReloadConfigString(rules("tcp"))
require.NoError(t, res.Err)
assert.False(t, res.Failed)
assert.NotSame(t, fw, f.firewall)
}
// Every input the firewall is built from rebuilds it when it changes, and nothing else does
func TestReloadFirewall_RebuildsOnAnyInput(t *testing.T) {
l := test.NewLogger()
networks := []netip.Prefix{netip.MustParsePrefix("10.0.0.1/24")}
base := "firewall:\n inbound:\n - port: any\n proto: any\n host: any\n"
setup := func() (*config.C, *Interface, *PKI) {
pki := &PKI{}
pki.cs.Store(&CertState{v2Cert: &dummyCert{version: cert.Version2, networks: networks}, initiatingVersion: cert.Version2})
cfg := config.NewC(l)
require.NoError(t, cfg.LoadString(base))
fw, err := NewFirewallFromConfig(l, pki.getCertState(), cfg)
require.NoError(t, err)
f := &Interface{pki: pki, firewall: fw, l: l}
cfg.RegisterReloadCallback(f.reloadFirewall)
return cfg, f, pki
}
setCert := func(pki *PKI, networks, unsafe []netip.Prefix) {
pki.cs.Store(&CertState{v2Cert: &dummyCert{version: cert.Version2, networks: networks, unsafeNetworks: unsafe}, initiatingVersion: cert.Version2})
}
cases := map[string]struct {
cfg string
cert func(*PKI)
want bool
}{
"nothing": {base + "punchy:\n punch: true\n", nil, false},
"rules": {base + " outbound:\n - port: any\n proto: any\n host: any\n", nil, true},
"inbound_action": {base + " inbound_action: reject\n", nil, true},
"conntrack": {base + " conntrack:\n tcp_timeout: 1m\n", nil, true},
"networks": {base, func(p *PKI) { setCert(p, []netip.Prefix{netip.MustParsePrefix("10.0.0.2/24")}, nil) }, true},
"unsafe_networks": {base, func(p *PKI) { setCert(p, networks, []netip.Prefix{netip.MustParsePrefix("192.168.0.0/24")}) }, true},
}
for name, tc := range cases {
t.Run(name, func(t *testing.T) {
cfg, f, pki := setup()
old := f.firewall
if tc.cert != nil {
tc.cert(pki)
}
res := cfg.ReloadConfigString(tc.cfg)
require.NoError(t, res.Err)
assert.False(t, res.Failed)
assert.Equal(t, tc.want, f.firewall != old)
})
}
}
+13
View File
@@ -32,10 +32,23 @@ const (
IPProtocolICMPv6 = 58 IPProtocolICMPv6 = 58
IPProtocolTCP = 6 IPProtocolTCP = 6
IPProtocolUDP = 17 IPProtocolUDP = 17
IPProtocolUDPLite = 136
IPProtocolDCCP = 33
IPProtocolSCTP = 132
ICMPv6TypeEchoRequest = 128 ICMPv6TypeEchoRequest = 128
ICMPv6TypeEchoReply = 129 ICMPv6TypeEchoReply = 129
) )
// HasPorts reports whether packets of proto carry port numbers that the packet parsers read: TCP, UDP, UDPLite,
// DCCP, and SCTP. Firewall rules with a port or a port range only apply to these protocols.
func HasPorts(proto uint8) bool {
switch proto {
case IPProtocolTCP, IPProtocolUDP, IPProtocolUDPLite, IPProtocolDCCP, IPProtocolSCTP:
return true
}
return false
}
func CreateRejectPacket(packet []byte, out []byte) []byte { func CreateRejectPacket(packet []byte, out []byte) []byte {
if len(packet) < 1 { if len(packet) < 1 {
return nil return nil
+9 -38
View File
@@ -15,7 +15,6 @@ import (
"time" "time"
"github.com/gaissmai/bart" "github.com/gaissmai/bart"
"github.com/rcrowley/go-metrics"
"github.com/slackhq/nebula/cert" "github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/header" "github.com/slackhq/nebula/header"
@@ -82,8 +81,6 @@ type LightHouse struct {
updateTrigger chan struct{} updateTrigger chan struct{}
queryChan chan netip.Addr queryChan chan netip.Addr
// queryDropped counts queries QueryServer dropped because queryChan was full.
queryDropped metrics.Counter
calculatedRemotes atomic.Pointer[bart.Table[[]*calculatedRemote]] // Maps VpnAddr to []*calculatedRemote calculatedRemotes atomic.Pointer[bart.Table[[]*calculatedRemote]] // Maps VpnAddr to []*calculatedRemote
@@ -119,8 +116,7 @@ func NewLightHouseFromConfig(ctx context.Context, l *slog.Logger, c *config.C, c
nebulaPort: nebulaPort, nebulaPort: nebulaPort,
punchy: p, punchy: p,
updateTrigger: make(chan struct{}, 1), updateTrigger: make(chan struct{}, 1),
queryChan: make(chan netip.Addr, c.GetUint32("handshakes.query_buffer", 1024)), queryChan: make(chan netip.Addr, c.GetUint32("handshakes.query_buffer", 64)),
queryDropped: metrics.GetOrRegisterCounter("lighthouse.query.dropped", nil),
l: l, l: l,
} }
h.localAddrsFn = func(al *LocalAllowList) []netip.Addr { h.localAddrsFn = func(al *LocalAllowList) []netip.Addr {
@@ -138,21 +134,18 @@ func NewLightHouseFromConfig(ctx context.Context, l *slog.Logger, c *config.C, c
h.metrics = newLighthouseMetrics() h.metrics = newLighthouseMetrics()
} }
err := h.reload(c, true, nil) err := h.reload(c, true)
if err != nil { if err != nil {
return nil, err return nil, err
} }
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) { c.RegisterReloadCallback(func(c *config.C) {
r.RestartRequiredIfChanged("lighthouse.am_lighthouse", "listen.port", "handshakes.query_buffer", "stats.lighthouse_metrics") err := h.reload(c, false)
err := h.reload(c, false, r)
switch v := err.(type) { switch v := err.(type) {
case *util.ContextualError: case *util.ContextualError:
v.Log(l) v.Log(l)
r.Failed()
case error: case error:
l.Error("failed to reload lighthouse", "error", err) l.Error("failed to reload lighthouse", "error", err)
r.Failed()
} }
}) })
@@ -193,8 +186,7 @@ func (lh *LightHouse) GetUpdateInterval() int64 {
return lh.interval.Load() return lh.interval.Load()
} }
// r is nil on the initial load func (lh *LightHouse) reload(c *config.C, initial bool) error {
func (lh *LightHouse) reload(c *config.C, initial bool, r *config.Reload) error {
if initial || c.HasChanged("lighthouse.advertise_addrs") { if initial || c.HasChanged("lighthouse.advertise_addrs") {
rawAdvAddrs := c.GetStringSlice("lighthouse.advertise_addrs", []string{}) rawAdvAddrs := c.GetStringSlice("lighthouse.advertise_addrs", []string{})
advAddrs := make([]netip.AddrPort, 0) advAddrs := make([]netip.AddrPort, 0)
@@ -229,7 +221,6 @@ func (lh *LightHouse) reload(c *config.C, initial bool, r *config.Reload) error
"addr", rawAddr, "addr", rawAddr,
"entry", i+1, "entry", i+1,
) )
r.Failed()
continue continue
} }
@@ -364,13 +355,7 @@ func (lh *LightHouse) reload(c *config.C, initial bool, r *config.Reload) error
} }
} }
// am_relay decides whether relay.relays applies to us at all if initial || c.HasChanged("relay.relays") {
if initial || c.HasChanged("relay.relays") || c.HasChanged("relay.am_relay") {
var before []netip.Addr
if !initial {
before = lh.GetRelaysForMe()
}
switch c.GetBool("relay.am_relay", false) { switch c.GetBool("relay.am_relay", false) {
case true: case true:
// Relays aren't allowed to specify other relays // Relays aren't allowed to specify other relays
@@ -388,7 +373,6 @@ func (lh *LightHouse) reload(c *config.C, initial bool, r *config.Reload) error
"relay", v, "relay", v,
"error", err, "error", err,
) )
r.Failed()
} else { } else {
lh.l.Info("Read relay from config", "relay", v) lh.l.Info("Read relay from config", "relay", v)
relaysForMe = append(relaysForMe, configRIP) relaysForMe = append(relaysForMe, configRIP)
@@ -396,11 +380,6 @@ func (lh *LightHouse) reload(c *config.C, initial bool, r *config.Reload) error
} }
lh.relaysForMe.Store(&relaysForMe) lh.relaysForMe.Store(&relaysForMe)
} }
// Tell the lighthouses now rather than at the next interval, but only when what we advertise changed
if !initial && !slices.Equal(before, lh.GetRelaysForMe()) {
lh.TriggerUpdate()
}
} }
return nil return nil
@@ -533,19 +512,14 @@ func (lh *LightHouse) Query(vpnAddr netip.Addr) *RemoteList {
return nil return nil
} }
// QueryServer is asynchronous and non-blocking, no reply should be expected. // QueryServer is asynchronous so no reply should be expected
// If the query channel is full, the query is dropped.
func (lh *LightHouse) QueryServer(vpnAddr netip.Addr) { func (lh *LightHouse) QueryServer(vpnAddr netip.Addr) {
// Don't put lighthouse addrs in the query channel because we can't query lighthouses about lighthouses // Don't put lighthouse addrs in the query channel because we can't query lighthouses about lighthouses
if lh.amLighthouse || lh.IsLighthouseAddr(vpnAddr) { if lh.amLighthouse || lh.IsLighthouseAddr(vpnAddr) {
return return
} }
select { lh.queryChan <- vpnAddr
case lh.queryChan <- vpnAddr:
default:
lh.queryDropped.Inc(1)
}
} }
func (lh *LightHouse) QueryCache(vpnAddrs []netip.Addr) *RemoteList { func (lh *LightHouse) QueryCache(vpnAddrs []netip.Addr) *RemoteList {
@@ -1261,13 +1235,10 @@ func (lhh *LightHouseHandler) handleHostQuery(n *NebulaMeta, fromVpnAddrs []neti
// sendHostPunchNotification signals the other side to punch some zero byte udp packets // sendHostPunchNotification signals the other side to punch some zero byte udp packets
func (lhh *LightHouseHandler) sendHostPunchNotification(n *NebulaMeta, fromVpnAddrs []netip.Addr, punchNotifDest netip.Addr, w EncWriter) { func (lhh *LightHouseHandler) sendHostPunchNotification(n *NebulaMeta, fromVpnAddrs []netip.Addr, punchNotifDest netip.Addr, w EncWriter) {
whereToPunch := fromVpnAddrs[0] whereToPunch := fromVpnAddrs[0]
// Resolve the target before taking the querier's RemoteList lock below. GetHostInfo takes the
// hostmap lock, and the hostmap is taken before the lighthouse, which is taken before a
// RemoteList, so looking it up inside the callback would deadlock.
targetHI := lhh.lh.ifce.GetHostInfo(punchNotifDest)
found, ln, err := lhh.lh.queryAndPrepMessage(whereToPunch, func(c *cache) (int, error) { found, ln, err := lhh.lh.queryAndPrepMessage(whereToPunch, func(c *cache) (int, error) {
n = lhh.resetMeta() n = lhh.resetMeta()
n.Type = NebulaMeta_HostPunchNotification n.Type = NebulaMeta_HostPunchNotification
targetHI := lhh.lh.ifce.GetHostInfo(punchNotifDest)
var useVersion cert.Version var useVersion cert.Version
if targetHI == nil { if targetHI == nil {
useVersion = lhh.lh.ifce.GetCertState().initiatingVersion useVersion = lhh.lh.ifce.GetCertState().initiatingVersion
+10 -257
View File
@@ -6,16 +6,13 @@ import (
"errors" "errors"
"fmt" "fmt"
"net/netip" "net/netip"
"sync"
"testing" "testing"
"time"
"github.com/gaissmai/bart" "github.com/gaissmai/bart"
"github.com/slackhq/nebula/cert" "github.com/slackhq/nebula/cert"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/header" "github.com/slackhq/nebula/header"
"github.com/slackhq/nebula/test" "github.com/slackhq/nebula/test"
"github.com/slackhq/nebula/udp"
"github.com/stretchr/testify/assert" "github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require" "github.com/stretchr/testify/require"
"go.yaml.in/yaml/v3" "go.yaml.in/yaml/v3"
@@ -87,15 +84,15 @@ func TestReloadLighthouseInterval(t *testing.T) {
lh.ifce = &mockEncWriter{} lh.ifce = &mockEncWriter{}
// The first one routine is kicked off by main.go currently, lets make sure that one dies // The first one routine is kicked off by main.go currently, lets make sure that one dies
require.NoError(t, c.ReloadConfigString("lighthouse:\n interval: 5").Err) require.NoError(t, c.ReloadConfigString("lighthouse:\n interval: 5"))
assert.Equal(t, int64(5), lh.interval.Load()) assert.Equal(t, int64(5), lh.interval.Load())
// Subsequent calls are killed off by the LightHouse.Reload function // Subsequent calls are killed off by the LightHouse.Reload function
require.NoError(t, c.ReloadConfigString("lighthouse:\n interval: 10").Err) require.NoError(t, c.ReloadConfigString("lighthouse:\n interval: 10"))
assert.Equal(t, int64(10), lh.interval.Load()) assert.Equal(t, int64(10), lh.interval.Load())
// If this completes then nothing is stealing our reload routine // If this completes then nothing is stealing our reload routine
require.NoError(t, c.ReloadConfigString("lighthouse:\n interval: 11").Err) require.NoError(t, c.ReloadConfigString("lighthouse:\n interval: 11"))
assert.Equal(t, int64(11), lh.interval.Load()) assert.Equal(t, int64(11), lh.interval.Load())
} }
@@ -296,7 +293,7 @@ func TestLighthouse_reload(t *testing.T) {
require.NoError(t, err) require.NoError(t, err)
c.ReloadConfigString(string(rc)) c.ReloadConfigString(string(rc))
err = lh.reload(c, false, nil) err = lh.reload(c, false)
require.NoError(t, err) require.NoError(t, err)
} }
@@ -334,7 +331,7 @@ func TestLighthouse_reloadStaticHostMap(t *testing.T) {
} }
rc, err := yaml.Marshal(nc) rc, err := yaml.Marshal(nc)
require.NoError(t, err) require.NoError(t, err)
require.NoError(t, c.ReloadConfigString(string(rc)).Err) require.NoError(t, c.ReloadConfigString(string(rc)))
rl := lh.Query(staticHost) rl := lh.Query(staticHost)
require.NotNil(t, rl) require.NotNil(t, rl)
@@ -350,7 +347,7 @@ func TestLighthouse_reloadStaticHostMap(t *testing.T) {
} }
rc, err = yaml.Marshal(nc) rc, err = yaml.Marshal(nc)
require.NoError(t, err) require.NoError(t, err)
require.NoError(t, c.ReloadConfigString(string(rc)).Err) require.NoError(t, c.ReloadConfigString(string(rc)))
rl = lh.Query(staticHost) rl = lh.Query(staticHost)
require.NotNil(t, rl) require.NotNil(t, rl)
@@ -358,7 +355,7 @@ func TestLighthouse_reloadStaticHostMap(t *testing.T) {
assert.Equal(t, []netip.AddrPort{netip.MustParseAddrPort("1.1.1.1:4242")}, rl.CopyAddrs([]netip.Prefix{})) assert.Equal(t, []netip.AddrPort{netip.MustParseAddrPort("1.1.1.1:4242")}, rl.CopyAddrs([]netip.Prefix{}))
// Reload with the same config. An unchanged entry must not duplicate. // Reload with the same config. An unchanged entry must not duplicate.
require.NoError(t, c.ReloadConfigString(string(rc)).Err) require.NoError(t, c.ReloadConfigString(string(rc)))
rl = lh.Query(staticHost) rl = lh.Query(staticHost)
require.NotNil(t, rl) require.NotNil(t, rl)
@@ -373,7 +370,7 @@ func TestLighthouse_reloadStaticHostMap(t *testing.T) {
} }
rc, err = yaml.Marshal(nc) rc, err = yaml.Marshal(nc)
require.NoError(t, err) require.NoError(t, err)
require.NoError(t, c.ReloadConfigString(string(rc)).Err) require.NoError(t, c.ReloadConfigString(string(rc)))
// Add a second host alongside the first. Both should be present, neither duplicated. // Add a second host alongside the first. Both should be present, neither duplicated.
nc = map[string]any{ nc = map[string]any{
@@ -384,7 +381,7 @@ func TestLighthouse_reloadStaticHostMap(t *testing.T) {
} }
rc, err = yaml.Marshal(nc) rc, err = yaml.Marshal(nc)
require.NoError(t, err) require.NoError(t, err)
require.NoError(t, c.ReloadConfigString(string(rc)).Err) require.NoError(t, c.ReloadConfigString(string(rc)))
rl = lh.Query(staticHost) rl = lh.Query(staticHost)
require.NotNil(t, rl) require.NotNil(t, rl)
@@ -406,7 +403,7 @@ func TestLighthouse_reloadStaticHostMap(t *testing.T) {
} }
rc, err = yaml.Marshal(nc) rc, err = yaml.Marshal(nc)
require.NoError(t, err) require.NoError(t, err)
require.NoError(t, c.ReloadConfigString(string(rc)).Err) require.NoError(t, c.ReloadConfigString(string(rc)))
_, isStatic := lh.GetStaticHostList()[staticHost] _, isStatic := lh.GetStaticHostList()[staticHost]
assert.False(t, isStatic) assert.False(t, isStatic)
@@ -798,247 +795,3 @@ func TestLightHouse_logLocalAddrsErr(t *testing.T) {
lh.logLocalAddrsErr(errors.New("something else")) lh.logLocalAddrsErr(errors.New("something else"))
assert.Contains(t, out.String(), "level=WARN") assert.Contains(t, out.String(), "level=WARN")
} }
// heldEncWriter holds the query worker in its first GetHostInfo until release closes.
type heldEncWriter struct {
mockEncWriter
once sync.Once
held chan struct{}
release chan struct{}
}
func (w *heldEncWriter) GetHostInfo(netip.Addr) *HostInfo {
w.once.Do(func() { close(w.held) })
<-w.release
return nil
}
// QueryServer never waits on the query worker, since the packet readers call it and nothing drains the channel at
// shutdown; a query that finds the channel full is dropped and counted. The worker is held on its first query so the
// channel fills.
func TestLighthouse_QueryServerDoesNotBlock(t *testing.T) {
l := test.NewLogger()
cs := testCertState(netip.MustParsePrefix("10.128.0.1/16"))
lh1 := "10.128.0.2"
c := config.NewC(l)
c.Settings["lighthouse"] = map[string]any{"hosts": []any{lh1}}
c.Settings["static_host_map"] = map[string]any{lh1: []any{"1.1.1.1:4242"}}
c.Settings["handshakes"] = map[string]any{"query_buffer": 1}
lh, err := NewLightHouseFromConfig(t.Context(), l, c, cs, nil, nil)
require.NoError(t, err)
w := &heldEncWriter{held: make(chan struct{}), release: make(chan struct{})}
lh.ifce = w
t.Cleanup(func() { close(w.release) })
lh.QueryServer(netip.MustParseAddr("10.128.0.10"))
select {
case <-w.held:
case <-time.After(5 * time.Second):
t.Fatal("the query worker never took the first query")
}
dropped := lh.queryDropped.Count()
done := make(chan struct{})
go func() {
lh.QueryServer(netip.MustParseAddr("10.128.0.11"))
lh.QueryServer(netip.MustParseAddr("10.128.0.12"))
close(done)
}()
select {
case <-done:
case <-time.After(5 * time.Second):
t.Fatal("QueryServer blocked on a full query channel")
}
assert.EqualValues(t, 1, lh.queryDropped.Count()-dropped)
}
// hostmapEncWriter answers GetHostInfo from a real hostmap, the way Interface does.
type hostmapEncWriter struct {
mockEncWriter
hm *HostMap
}
func (w *hostmapEncWriter) GetHostInfo(a netip.Addr) *HostInfo { return w.hm.QueryVpnAddr(a) }
func (w *hostmapEncWriter) GetCertState() *CertState {
return &CertState{initiatingVersion: cert.Version1}
}
// A lighthouse runs four paths that each hold one lock while waiting for the next: CheckAndComplete
// holds the hostmap and takes the handshake manager, StartHandshake holds the handshake manager and
// takes the lighthouse, a host update holds the lighthouse and takes the sender's RemoteList, and a
// punch notification holds the querier's RemoteList while it resolves the target. If that last step
// reaches the hostmap under the RemoteList lock, the four deadlock. Drive them all at once.
func TestLighthouse_PunchNotificationDoesNotCloseLockCycle(t *testing.T) {
l := test.NewLogger()
myVpnNet := netip.MustParsePrefix("10.128.0.1/24")
c := config.NewC(l)
c.Settings["lighthouse"] = map[string]any{"am_lighthouse": true}
c.Settings["listen"] = map[string]any{"port": 4242}
lh, err := NewLightHouseFromConfig(t.Context(), l, c, testCertState(myVpnNet), nil, nil)
require.NoError(t, err)
mainHM := newHostMap(l)
pr := []netip.Prefix{}
mainHM.preferredRanges.Store(&pr)
lh.ifce = &hostmapEncWriter{hm: mainHM}
hm := NewHandshakeManager(l, mainHM, lh, &udp.NoopConn{}, defaultHandshakeConfig)
f := &Interface{hostMap: mainHM, lightHouse: lh, handshakeManager: hm, l: l}
hm.f = f
peer := netip.MustParseAddr("10.128.0.2")
peerUdp := netip.MustParseAddrPort("10.0.0.2:4242")
target := netip.MustParseAddr("10.128.0.3")
targetUdp := netip.MustParseAddrPort("10.0.0.3:4242")
outbound := netip.MustParseAddr("10.128.0.4")
inbound := netip.MustParseAddr("10.128.0.5")
seed := lh.NewRequestHandler()
newLHHostUpdate(peerUdp, peer, []netip.AddrPort{peerUdp}, seed)
newLHHostUpdate(targetUdp, target, []netip.AddrPort{targetUdp}, seed)
stop := make(chan struct{})
var wg sync.WaitGroup
loop := func(body func()) {
defer wg.Done()
for {
select {
case <-stop:
return
default:
body()
}
}
}
wg.Add(4)
// Outbound handshakes keep starting and being dropped.
go loop(func() {
hm.DeleteHostInfo(hm.StartHandshake(outbound, nil))
})
// Inbound handshakes keep completing and being torn down.
var idx uint32
go loop(func() {
idx++
hi := &HostInfo{
vpnAddrs: []netip.Addr{inbound},
localIndexId: idx,
remoteIndexId: idx,
remotes: lh.QueryCache([]netip.Addr{inbound}),
HandshakePacket: map[uint8][]byte{handshakePacketStage0: {byte(idx)}},
}
if _, err := hm.CheckAndComplete(hi, handshakePacketStage0, f); err == nil {
mainHM.DeleteHostInfo(hi)
}
})
// The peer keeps sending host updates.
updates := lh.NewRequestHandler()
go loop(func() {
newLHHostUpdate(peerUdp, peer, []netip.AddrPort{peerUdp}, updates)
})
// The same peer keeps querying for the target, which ends in a punch notification.
queries := lh.NewRequestHandler()
go loop(func() {
sendLHHostRequest(peerUdp, peer, target, queries, nil)
})
time.Sleep(time.Second)
close(stop)
done := make(chan struct{})
go func() { wg.Wait(); close(done) }()
select {
case <-done:
case <-time.After(5 * time.Second):
t.Fatal("deadlock: the lighthouse paths never finished after stop was signalled")
}
}
func TestLighthouse_reloadRelaysTriggersUpdate(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
base := "lighthouse:\n am_lighthouse: true\nlisten:\n port: 4242\n"
require.NoError(t, c.LoadString(base+"relay:\n relays: [10.128.0.5]\n"))
lh, err := NewLightHouseFromConfig(t.Context(), l, c, testCertState(netip.MustParsePrefix("10.128.0.1/24")), nil, nil)
require.NoError(t, err)
assert.Empty(t, lh.updateTrigger)
require.NoError(t, c.ReloadConfigString(base+"relay:\n relays: [10.128.0.6]\n").Err)
assert.Equal(t, []netip.Addr{netip.MustParseAddr("10.128.0.6")}, lh.GetRelaysForMe())
require.Len(t, lh.updateTrigger, 1)
<-lh.updateTrigger
// The config changed but what we advertise did not, a bad entry is skipped
require.NoError(t, c.ReloadConfigString(base+"relay:\n relays: [10.128.0.6, nope]\n").Err)
assert.Equal(t, []netip.Addr{netip.MustParseAddr("10.128.0.6")}, lh.GetRelaysForMe())
assert.Empty(t, lh.updateTrigger)
}
func TestLighthouse_reloadAmRelayAlone(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
base := "lighthouse:\n am_lighthouse: true\nlisten:\n port: 4242\nrelay:\n relays: [10.128.0.5]\n"
require.NoError(t, c.LoadString(base+" am_relay: false\n"))
lh, err := NewLightHouseFromConfig(t.Context(), l, c, testCertState(netip.MustParsePrefix("10.128.0.1/24")), nil, nil)
require.NoError(t, err)
assert.Equal(t, []netip.Addr{netip.MustParseAddr("10.128.0.5")}, lh.GetRelaysForMe())
// Relays aren't allowed to have relays, flipping am_relay alone must drop them and tell the lighthouses
require.NoError(t, c.ReloadConfigString(base+" am_relay: true\n").Err)
assert.Empty(t, lh.GetRelaysForMe())
require.Len(t, lh.updateTrigger, 1)
<-lh.updateTrigger
require.NoError(t, c.ReloadConfigString(base+" am_relay: false\n").Err)
assert.Equal(t, []netip.Addr{netip.MustParseAddr("10.128.0.5")}, lh.GetRelaysForMe())
require.Len(t, lh.updateTrigger, 1)
}
func TestLighthouse_reloadRestartRequired(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
base := "lighthouse:\n am_lighthouse: true\n"
require.NoError(t, c.LoadString(base+"listen:\n port: 4242\n"))
_, err := NewLightHouseFromConfig(t.Context(), l, c, testCertState(netip.MustParsePrefix("10.128.0.1/24")), nil, nil)
require.NoError(t, err)
res := c.ReloadConfigString(base + "listen:\n port: 4242\n")
require.NoError(t, res.Err)
assert.False(t, res.RestartRequired)
res = c.ReloadConfigString(base + "listen:\n port: 4243\n")
require.NoError(t, res.Err)
assert.True(t, res.RestartRequired)
}
// A relay that doesn't parse is skipped and the rest are used, the reload says one was skipped
func TestLighthouse_reloadBadRelayFails(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
require.NoError(t, c.LoadString("listen:\n port: 4242\n"))
lh, err := NewLightHouseFromConfig(t.Context(), l, c, testCertState(netip.MustParsePrefix("10.128.0.1/24")), nil, nil)
require.NoError(t, err)
res := c.ReloadConfigString("listen:\n port: 4242\nrelay:\n relays: [10.128.0.2, nope]\n")
require.NoError(t, res.Err)
assert.True(t, res.Failed)
assert.Equal(t, []netip.Addr{netip.MustParseAddr("10.128.0.2")}, lh.GetRelaysForMe())
}
// An advertise_addrs entry inside our own network is skipped and the rest are used, the reload says one was skipped
func TestLighthouse_reloadAdvertiseAddrInNetworkFails(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
require.NoError(t, c.LoadString("listen:\n port: 4242\n"))
lh, err := NewLightHouseFromConfig(t.Context(), l, c, testCertState(netip.MustParsePrefix("10.128.0.1/24")), nil, nil)
require.NoError(t, err)
res := c.ReloadConfigString("listen:\n port: 4242\nlighthouse:\n advertise_addrs: [\"1.2.3.4:4242\", \"10.128.0.5:4242\"]\n")
require.NoError(t, res.Err)
assert.True(t, res.Failed)
assert.Equal(t, []netip.AddrPort{netip.MustParseAddrPort("1.2.3.4:4242")}, lh.GetAdvertiseAddrs())
}
-79
View File
@@ -1,79 +0,0 @@
package localapi
import (
"errors"
"fmt"
"os"
)
// Access is who may reach a socket beyond the account nebula runs as: the node's mode,
// and the group that shares the socket when the config names one.
//
// It is the dockerd model (-G): the operator names a group, nebula hands the socket to
// it, and who is in the group is managed outside nebula. On unix the mode is the
// socket node's and the group becomes its owning group; on windows the group is an
// entry in the pipe's security descriptor and the mode means nothing to a pipe.
type Access struct {
Mode os.FileMode
Group Group
}
// Group is a group named in the config, resolved to what the platform checks. The zero
// Group is none.
type Group struct {
// Name is what the config said.
Name string
// ID is the platform's identifier for it: a gid on unix, a SID on windows.
ID string
}
const (
// ownerOnlyMode is a socket's mode when no group shares it: the account nebula runs
// as, and nobody else.
ownerOnlyMode os.FileMode = 0600
// groupSocketMode is the mode a socket takes when a group shares it and the config
// names no other: the account nebula runs as and the group, and nobody else.
groupSocketMode os.FileMode = 0660
)
// errReopenForAccess is returned by applyAccess on a platform whose access control is
// fixed when the endpoint is created, so a change means creating it again.
var errReopenForAccess = errors.New("access is fixed at listen time")
// ownerOnly is the control socket's access when no group is configured. The mode
// itself is not configurable there; what an operator can choose is a group,
// CtlGroupKey, which makes it groupSocketMode.
func ownerOnly() Access { return Access{Mode: ownerOnlyMode} }
// shared reports whether a group shares the socket.
func (a Access) shared() bool { return a.Group.ID != "" }
// String renders the access for a log line.
func (a Access) String() string {
if !a.shared() {
return modeString(a.Mode)
}
return fmt.Sprintf("%s group %s (%s)", modeString(a.Mode), a.Group.Name, a.Group.ID)
}
// groupFromConfig reads the group a key names, the zero Group when it names none. The
// name is resolved here, when the config is read, so a group that does not exist is a
// config error reported with the key rather than a bind that fails later.
func groupFromConfig(c interface{ GetString(string, string) string }, key string) (Group, error) {
name := c.GetString(key, "")
if name == "" {
return Group{}, nil
}
id, err := lookupGroupID(name)
if err != nil {
return Group{}, fmt.Errorf("%s: %w", key, err)
}
return Group{Name: name, ID: id}, nil
}
func modeString(m os.FileMode) string {
return fmt.Sprintf("%04o", m)
}
-176
View File
@@ -1,176 +0,0 @@
//go:build !windows
package localapi
import (
"context"
"os"
"os/user"
"strconv"
"syscall"
"testing"
"github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"golang.org/x/sys/unix"
)
// ownGroup returns a group this process may hand a socket to: its own effective group,
// by name and by gid, which is the one chown accepts from an unprivileged user.
func ownGroup(t *testing.T) (name, gid string) {
t.Helper()
gid = strconv.Itoa(os.Getegid())
g, err := user.LookupGroupId(gid)
require.NoError(t, err)
return g.Name, gid
}
func socketGroup(t *testing.T, path string) uint32 {
t.Helper()
fi, err := os.Lstat(path)
require.NoError(t, err)
st, ok := fi.Sys().(*syscall.Stat_t)
require.True(t, ok)
return uint32(st.Gid)
}
// TestLookupGroupID: a group is named as the host names it or by its gid; a gid the
// host does not list is accepted as chown would accept it, and a name it does not list
// is a mistake to report.
func TestLookupGroupID(t *testing.T) {
name, gid := ownGroup(t)
got, err := lookupGroupID(name)
require.NoError(t, err)
assert.Equal(t, gid, got)
got, err = lookupGroupID(gid)
require.NoError(t, err)
assert.Equal(t, gid, got)
got, err = lookupGroupID("4000000000")
require.NoError(t, err)
assert.Equal(t, "4000000000", got)
_, err = lookupGroupID("no-such-group-nebula-test")
require.ErrorContains(t, err, "does not exist")
}
// TestControlAccessFromConfig: no group is the owner-only socket; a group is 0660 with
// that group; a group the host does not know is a config error naming the key.
func TestControlAccessFromConfig(t *testing.T) {
name, gid := ownGroup(t)
load := func(body string) (Access, error) {
c := config.NewC(test.NewLogger())
require.NoError(t, c.LoadString(body))
return controlAccessFrom(c)
}
acc, err := load("local_api: {}\n")
require.NoError(t, err)
assert.Equal(t, ownerOnly(), acc)
acc, err = load("local_api:\n control:\n group: " + name + "\n")
require.NoError(t, err)
assert.Equal(t, os.FileMode(0660), acc.Mode)
assert.Equal(t, name, acc.Group.Name)
assert.Equal(t, gid, acc.Group.ID)
_, err = load("local_api:\n control:\n group: no-such-group-nebula-test\n")
require.ErrorContains(t, err, CtlGroupKey)
assert.ErrorContains(t, err, "does not exist")
}
// TestControlSocketIsHandedToTheGroup: the socket is created with the group's gid and
// 0660, whatever the umask, and a reload that takes the group away tightens the socket
// in place, without moving it.
func TestControlSocketIsHandedToTheGroup(t *testing.T) {
old := unix.Umask(0077)
defer unix.Umask(old)
name, gid := ownGroup(t)
want, err := strconv.ParseUint(gid, 10, 32)
require.NoError(t, err)
s := NewServer(test.NewLogger(), basicRegistry())
path := testSocket(t)
require.NoError(t, s.Rebind(context.Background(), path, Access{Mode: 0660, Group: Group{Name: name, ID: gid}}))
defer s.Stop()
fi, err := os.Lstat(path)
require.NoError(t, err)
assert.Equal(t, os.FileMode(0660), fi.Mode().Perm())
assert.Equal(t, uint32(want), socketGroup(t, path))
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
fi, err = os.Lstat(path)
require.NoError(t, err)
assert.Equal(t, os.FileMode(0600), fi.Mode().Perm())
assert.Equal(t, path, s.Path())
_, err = (&Client{Path: path}).Describe()
require.NoError(t, err, "the same listener is still serving")
}
// TestControlSocketWithAGroupTrustsTheKernel: with a group sharing the socket, the
// kernel's check on connect is the access control and the daemon's peer check stands
// aside, so a connection the peer check alone would refuse is served. Without a group
// the same connection is refused; see TestPeerCheckRefusesAnotherUser.
func TestControlSocketWithAGroupTrustsTheKernel(t *testing.T) {
if !peerCheckSupported {
t.Skip("no peer credentials on this platform")
}
name, gid := ownGroup(t)
me := currentUID()
asUID(t, me+1)
s := NewServer(test.NewLogger(), basicRegistry())
asUID(t, me)
path := testSocket(t)
require.NoError(t, s.Rebind(context.Background(), path, Access{Mode: 0660, Group: Group{Name: name, ID: gid}}))
defer s.Stop()
_, err := (&Client{Path: path}).Describe()
require.NoError(t, err, "the socket's group admitted us; the peer check must not refuse")
// Taking the group away on a reload puts the peer check back.
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
_, err = (&Client{Path: path}).Describe()
require.ErrorContains(t, err, "not permitted")
}
// TestRemovingTheGroupHandsTheSocketBack: a reload that drops the group key leaves the
// mode the operator set but returns the socket to the daemon's own group, so the old
// group's members are no longer admitted. The socket is first handed to a group this
// process belongs to other than its own, which not every host offers.
func TestRemovingTheGroupHandsTheSocketBack(t *testing.T) {
own := os.Getegid()
groups, err := os.Getgroups()
require.NoError(t, err)
other := -1
for _, g := range groups {
if g != own {
other = g
break
}
}
if other < 0 {
t.Skip("this process belongs to no group but its own")
}
gid := strconv.Itoa(other)
s := NewServer(test.NewLogger(), basicRegistry())
path := testSocket(t)
require.NoError(t, s.Rebind(context.Background(), path, Access{Mode: 0660, Group: Group{Name: gid, ID: gid}}))
defer s.Stop()
assert.Equal(t, uint32(other), socketGroup(t, path))
require.NoError(t, s.Rebind(context.Background(), path, Access{Mode: 0660}))
assert.Equal(t, uint32(own), socketGroup(t, path), "the daemon's own group again")
fi, err := os.Lstat(path)
require.NoError(t, err)
assert.Equal(t, os.FileMode(0660), fi.Mode().Perm(), "the mode is the operator's")
}
-44
View File
@@ -1,44 +0,0 @@
//go:build windows
package localapi
import (
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"golang.org/x/sys/windows"
)
// TestLookupGroupIDOnWindows: an account name resolves to its SID, a SID written out
// is accepted as it is, and an account the host does not know is reported.
func TestLookupGroupIDOnWindows(t *testing.T) {
users, err := windows.CreateWellKnownSid(windows.WinBuiltinUsersSid)
require.NoError(t, err)
got, err := lookupGroupID("Users")
require.NoError(t, err)
assert.Equal(t, users.String(), got)
got, err = lookupGroupID(users.String())
require.NoError(t, err)
assert.Equal(t, users.String(), got)
_, err = lookupGroupID("no-such-account-nebula-test")
require.ErrorContains(t, err, "does not exist")
}
// TestSDDLAdmitsTheGroup: the descriptor is the owner, SYSTEM and Administrators with
// full access, plus the group with what a client needs and nothing generic, so a
// member cannot create an instance of the pipe; see pipeClientAccess.
func TestSDDLAdmitsTheGroup(t *testing.T) {
sddl, err := sddlFor(ownerOnly())
require.NoError(t, err)
assert.NotContains(t, sddl, "S-1-5-32-545")
sddl, err = sddlFor(Access{Mode: groupSocketMode, Group: Group{Name: "Users", ID: "S-1-5-32-545"}})
require.NoError(t, err)
assert.Contains(t, sddl, "(A;;GA;;;SY)(A;;GA;;;BA)(A;;0x120083;;;S-1-5-32-545)")
assert.Zero(t, pipeClientAccess&windows.FILE_APPEND_DATA, "FILE_APPEND_DATA is FILE_CREATE_PIPE_INSTANCE on a pipe")
assert.Zero(t, pipeClientAccess&(windows.GENERIC_READ|windows.GENERIC_WRITE|windows.GENERIC_ALL))
}
-163
View File
@@ -1,163 +0,0 @@
package localapi
import (
"encoding/json"
"errors"
"fmt"
"io"
"net"
"time"
"github.com/slackhq/nebula/commands"
)
// DialTimeout bounds the connect to a socket that exists but is not being served.
const DialTimeout = 5 * time.Second
// ResponseTimeout bounds the exchange once connected: sending the request and waiting
// for the daemon's answer. The daemon drops a client that is slow to send in 10s and one
// that is slow to read in 30s, so this is the other side of the same bargain, with room
// for the slowest command (a hostmap dump on a large network) before a stalled daemon is
// given up on. Client.Timeout overrides it.
const ResponseTimeout = 60 * time.Second
// MaxResponseBytes bounds what a client will read back. A response carries one command's
// output, and a hostmap dump on a large network is the realistic worst case.
const MaxResponseBytes = 64 << 20
// Client talks to a nebula listening at Path.
type Client struct {
Path string
// Timeout bounds the request and response once connected. Zero means
// ResponseTimeout.
Timeout time.Duration
}
// ReplyError is an answer the client could not read: the daemon was reached and wrote
// something back, but not a response this client understands. The daemon is not
// nebula, or not a version this client speaks to.
type ReplyError struct {
Err error
}
func (e *ReplyError) Error() string { return "the daemon's reply could not be read: " + e.Err.Error() }
func (e *ReplyError) Unwrap() error { return e.Err }
// DaemonError is a refusal the daemon answered with: the connection worked, the request
// was read, and the daemon said no. It is told apart from a failure to reach the daemon
// so that a client does not send its user after a socket path, or a sudo, when the daemon
// was right there and answered. A protocol version the daemon does not speak, a socket at
// its connection limit and a peer the daemon will not serve all arrive as one of these.
type DaemonError struct {
Msg string
}
func (e *DaemonError) Error() string { return e.Msg }
// Describe asks the daemon which commands it serves and what parameters they take, so a
// client can offer them without having been built against this daemon.
func (c *Client) Describe() ([]commands.CommandSchema, error) {
res, err := c.do(&Request{Version: ProtocolVersion, Describe: true})
if err != nil {
return nil, err
}
if res.Error != "" {
return nil, &DaemonError{Msg: res.Error}
}
schema := []commands.CommandSchema{}
if err := json.Unmarshal(res.Result, &schema); err != nil {
return nil, &ReplyError{Err: fmt.Errorf("could not parse the command schema: %w", err)}
}
return schema, nil
}
// Invoke runs one command and returns what the daemon said. params are the command's
// options by name, as the strings a command line parsed them from; a program speaking
// the protocol directly may send JSON's own types instead, see Request.Params. An empty
// format asks for JSON, which is the API's default; "text" asks the daemon to apply the
// command's own human rendering.
func (c *Client) Invoke(command string, params map[string]string, args []string, format string) (*Response, error) {
var wire map[string]any
if len(params) > 0 {
wire = make(map[string]any, len(params))
for k, v := range params {
wire[k] = v
}
}
return c.do(&Request{
Version: ProtocolVersion,
Command: command,
Params: wire,
Args: args,
Format: format,
})
}
// do sends one request and reads the one response.
//
// Nothing is checked about the socket before the connect: its mode is the kernel's to
// enforce, and a socket that is not ours to connect to is refused there with a
// permission error, which the caller turns into advice.
func (c *Client) do(req *Request) (*Response, error) {
conn, err := dial(c.Path, DialTimeout)
if err != nil {
return nil, err
}
defer conn.Close()
// Without this a daemon that accepts and then stalls, blocked inside a command or
// stopped with SIGSTOP, hangs the client for good, and a script wrapping it never
// completes. One deadline covers the write and the read: the answer is due within
// the budget, however the time is spent.
timeout := c.Timeout
if timeout <= 0 {
timeout = ResponseTimeout
}
_ = conn.SetDeadline(time.Now().Add(timeout))
// If we get a write error, save that fact but then look for a response from the server.
// Some error conditions manifest as the server accepting our connection, writing an
// error, and then closing the connection. Our write races against that close.
writeErr := error(nil)
if err := writeFrame(conn, req); err != nil {
writeErr = fmt.Errorf("could not send request: %w", err)
}
res, err := readResponse(conn)
if err != nil {
// Something answered and it was not a response: the daemon was reached, and
// that is the message even when it hung up before the request was written,
// which a daemon that is not nebula is likely to do.
var fe *frameError
if errors.As(err, &fe) {
return nil, &ReplyError{Err: err}
}
if writeErr != nil {
return nil, writeErr
}
var ne net.Error
if errors.As(err, &ne) && ne.Timeout() {
return nil, fmt.Errorf("the daemon did not answer within %s", timeout)
}
return nil, err
}
return res, nil
}
func readResponse(conn net.Conn) (*Response, error) {
res := &Response{}
if err := readFrame(conn, MaxResponseBytes, "response", res, false); err != nil {
if errors.Is(err, io.EOF) {
return nil, errors.New("the daemon closed the connection without answering")
}
return nil, err
}
return res, nil
}
-39
View File
@@ -1,39 +0,0 @@
//go:build !windows
package localapi
import (
"fmt"
"os"
"os/user"
"strconv"
)
// lookupGroupID resolves a group's name, or a numeric gid, to the gid as a string. A
// gid that the group database does not list is accepted as chown would accept it: the
// operator may be naming a group the host will learn about later, or one that only a
// directory service knows.
func lookupGroupID(name string) (string, error) {
if g, err := user.LookupGroup(name); err == nil {
return g.Gid, nil
}
if _, err := strconv.ParseUint(name, 10, 32); err == nil {
return name, nil
}
return "", fmt.Errorf("group %q does not exist on this host", name)
}
// gidFor returns the gid the socket is owned by under acc: the configured group's, or
// the daemon's own effective group when none is configured.
func gidFor(acc Access) (int, error) {
if !acc.shared() {
return os.Getegid(), nil
}
gid, err := strconv.ParseUint(acc.Group.ID, 10, 32)
if err != nil {
return 0, fmt.Errorf("group %q resolved to %q, which is not a gid", acc.Group.Name, acc.Group.ID)
}
return int(gid), nil
}
-30
View File
@@ -1,30 +0,0 @@
//go:build windows
package localapi
import (
"fmt"
"strings"
"golang.org/x/sys/windows"
)
// lookupGroupID resolves an account name, local or domain qualified, or a SID written
// out, to the SID as a string, which is how the pipe's security descriptor names it.
// Any account the host can resolve is accepted, as docker accepts any SID for its -G: a
// group is the natural thing to name, but the descriptor does not care.
func lookupGroupID(name string) (string, error) {
if strings.HasPrefix(strings.ToUpper(name), "S-1-") {
sid, err := windows.StringToSid(name)
if err != nil {
return "", fmt.Errorf("%q is not a SID: %w", name, err)
}
return sid.String(), nil
}
sid, _, _, err := windows.LookupSID("", name)
if err != nil {
return "", fmt.Errorf("account %q does not exist on this host: %w", name, err)
}
return sid.String(), nil
}
-58
View File
@@ -1,58 +0,0 @@
package localapi
// DirName is the directory under /var/run that holds nebula's sockets by default.
const DirName = "nebula"
// CtlEnabledKey is the optional config key that turns the control socket off. It is on
// unless the config says false, and reloadable: a reload that turns it off removes the
// socket, and one that turns it back on brings it up again.
const CtlEnabledKey = "local_api.control.enabled"
// ControlEnabled reports whether a config leaves the control socket on. nebula ctl reads
// it too, so a client pointed at a config that turned the socket off says so rather than
// failing to dial.
func ControlEnabled(c interface{ GetBool(string, bool) bool }) bool {
return c.GetBool(CtlEnabledKey, true)
}
// CtlSocketKey is the optional config key naming the control socket. Unset, the socket
// is DefaultSocketPath, one fixed path per platform that both the daemon and nebula ctl
// know, which is what lets the endpoint work with no configuration on either side. It
// suits a daemon running as root, which is what the tun device needs anyway. A daemon run as another user, or a second daemon on
// one host, names its own. The CLI reads the same key, so when both sides see the same
// config file they agree on the socket with nothing more said.
const CtlSocketKey = "local_api.control.socket"
// CtlGroupKey is the optional config key naming a group that shares the control socket.
// Unset, the socket admits the account nebula runs as and nobody else. Set, the socket
// is handed to the group, 0660 on unix and an entry in the pipe's security descriptor on
// windows; see Access and Server.trustSocket.
const CtlGroupKey = "local_api.control.group"
// controlAccessFrom returns the control socket's access a config implies: the group
// CtlGroupKey names, if any, and the mode that goes with having one or not.
func controlAccessFrom(c interface{ GetString(string, string) string }) (Access, error) {
group, err := groupFromConfig(c, CtlGroupKey)
if err != nil {
return Access{}, err
}
if group.ID == "" {
return ownerOnly(), nil
}
return Access{Mode: groupSocketMode, Group: group}, nil
}
// SocketPathFromConfig returns the control socket path a config implies: CtlSocketKey
// when it is set, else DefaultSocketPath. A configured path must be absolute, so that
// the daemon and nebula ctl, started from wherever the operator happens to be, name the
// same socket from one config, and one this platform can bind; see checkSocketPath.
func SocketPathFromConfig(c interface{ GetString(string, string) string }) (string, error) {
override := c.GetString(CtlSocketKey, "")
if override == "" {
return DefaultSocketPath, nil
}
if err := checkSocketPath(CtlSocketKey, override); err != nil {
return "", err
}
return override, nil
}
-52
View File
@@ -1,52 +0,0 @@
//go:build !windows
package localapi
import (
"strings"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
type fakeConfig map[string]string
func (f fakeConfig) GetString(k, d string) string {
if v, ok := f[k]; ok {
return v
}
return d
}
// TestSocketPathFromConfigUsesTheDefault is what makes the endpoint work with no
// configuration: the daemon and the CLI both land on the one fixed path.
func TestSocketPathFromConfigUsesTheDefault(t *testing.T) {
p, err := SocketPathFromConfig(fakeConfig{})
require.NoError(t, err)
assert.Equal(t, "/var/run/nebula/ctl.sock", p)
assert.Equal(t, DefaultSocketPath, p)
}
func TestSocketPathFromConfigHonoursOverride(t *testing.T) {
p, err := SocketPathFromConfig(fakeConfig{CtlSocketKey: "/tmp/custom.sock"})
require.NoError(t, err)
assert.Equal(t, "/tmp/custom.sock", p, "the path is used as given")
}
// TestSocketPathFromConfigRefusesARelativeOverride: the daemon and the CLI are started
// from different directories, so a relative path would name two sockets from one config.
// It is refused rather than resolved against anything.
func TestSocketPathFromConfigRefusesARelativeOverride(t *testing.T) {
for _, rel := range []string{"run/ctl.sock", "./ctl.sock", "ctl.sock", "../ctl.sock"} {
_, err := SocketPathFromConfig(fakeConfig{CtlSocketKey: rel})
require.Error(t, err, rel)
require.ErrorContains(t, err, "must be an absolute path", rel)
assert.ErrorContains(t, err, CtlSocketKey, rel)
}
}
func TestSocketPathFromConfigRejectsOverlongOverride(t *testing.T) {
_, err := SocketPathFromConfig(fakeConfig{CtlSocketKey: "/" + strings.Repeat("x", sunPathMax)})
assert.ErrorContains(t, err, "too long for this platform")
}
-25
View File
@@ -1,25 +0,0 @@
//go:build darwin || freebsd
package localapi
import (
"net"
"golang.org/x/sys/unix"
)
// peerCheckSupported reports that this platform can identify a connecting process.
const peerCheckSupported = true
// peerUID returns the uid of the process on the other end of conn.
func peerUID(conn net.Conn) (uid uint32, err error) {
err = withPeerFD(conn, func(fd int) error {
xu, credErr := unix.GetsockoptXucred(fd, unix.SOL_LOCAL, unix.LOCAL_PEERCRED)
if credErr != nil {
return credErr
}
uid = xu.Uid
return nil
})
return uid, err
}
-25
View File
@@ -1,25 +0,0 @@
//go:build linux
package localapi
import (
"net"
"golang.org/x/sys/unix"
)
// peerCheckSupported reports that this platform can identify a connecting process.
const peerCheckSupported = true
// peerUID returns the uid of the process on the other end of conn.
func peerUID(conn net.Conn) (uid uint32, err error) {
err = withPeerFD(conn, func(fd int) error {
cred, credErr := unix.GetsockoptUcred(fd, unix.SOL_SOCKET, unix.SO_PEERCRED)
if credErr != nil {
return credErr
}
uid = cred.Uid
return nil
})
return uid, err
}
-62
View File
@@ -1,62 +0,0 @@
//go:build netbsd
package localapi
import (
"encoding/binary"
"fmt"
"net"
"unsafe"
"golang.org/x/sys/unix"
)
const peerCheckSupported = true
// NetBSD reports the peer's identity through LOCAL_PEEREID at the SOL_LOCAL level,
// filling a `struct unpcbid { pid_t unp_pid; uid_t unp_euid; gid_t unp_egid; }`.
// golang.org/x/sys/unix has no typed wrapper for it, so the getsockopt goes out by hand
// exactly the way that package's own generated wrappers do on this platform.
const (
solLocal = 0
localPeerEID = 0x0003
unpcbidSize = 12
unpcbidEUIDOffs = 4
)
// peerUID returns the uid of the process on the other end of conn.
//
// Anything unexpected here is an error rather than a default, so a wrong constant or an
// unsupported kernel refuses connections instead of silently admitting them.
func peerUID(conn net.Conn) (uid uint32, err error) {
err = withPeerFD(conn, func(fd int) error {
var (
buf [unpcbidSize]byte
size = uint32(unpcbidSize)
)
_, _, errno := unix.Syscall6(
unix.SYS_GETSOCKOPT,
uintptr(fd),
uintptr(solLocal),
uintptr(localPeerEID),
uintptr(unsafe.Pointer(&buf[0])),
uintptr(unsafe.Pointer(&size)),
0,
)
if errno != 0 {
return errno
}
if size != unpcbidSize {
return fmt.Errorf("LOCAL_PEEREID returned %d bytes, expected %d", size, unpcbidSize)
}
// Read, not cast: the buffer is bytes, and a cast would fault on an
// architecture that wants the uint32 aligned.
uid = binary.NativeEndian.Uint32(buf[unpcbidEUIDOffs:])
return nil
})
return uid, err
}
-34
View File
@@ -1,34 +0,0 @@
//go:build openbsd
package localapi
import (
"net"
"golang.org/x/sys/unix"
)
// peerCheckSupported reports that this platform can identify a connecting process.
const peerCheckSupported = true
// peerUID returns the uid of the process on the other end of conn.
//
// OpenBSD reports the peer through SO_PEERCRED at SOL_SOCKET, filling a
// `struct sockpeercred { uid_t uid; gid_t gid; pid_t pid; }`. System calls here have to
// go through libc, so the raw getsockopt netbsd uses is not an option, and
// golang.org/x/sys/unix has no typed wrapper for this struct. It does export
// GetsockoptInt, which goes through its libc trampoline with a 4 byte buffer, and the
// kernel copies out min(caller's length, option length): the first four bytes, which are
// the uid. That is a dependence on the field order, the same one netbsd's offset carries.
func peerUID(conn net.Conn) (uid uint32, err error) {
err = withPeerFD(conn, func(fd int) error {
v, credErr := unix.GetsockoptInt(fd, unix.SOL_SOCKET, unix.SO_PEERCRED)
if credErr != nil {
return credErr
}
// A uid_t is 32 bits; GetsockoptInt read exactly that many and widened it.
uid = uint32(v)
return nil
})
return uid, err
}
-20
View File
@@ -1,20 +0,0 @@
//go:build windows
package localapi
import (
"errors"
"net"
)
// peerCheckSupported is false where the connecting process is not identified separately.
//
// On windows there is nothing to add: the pipe's security descriptor already admits
// only the account nebula runs as, which is the check, not a second opinion on it.
const peerCheckSupported = false
var errNoPeerCheck = errors.New("peer credentials are not available on this platform")
func peerUID(conn net.Conn) (uint32, error) {
return 0, errNoPeerCheck
}
-302
View File
@@ -1,302 +0,0 @@
//go:build !windows
package localapi
import (
"errors"
"fmt"
"net"
"os"
"os/user"
"path/filepath"
"strconv"
"sync"
"syscall"
"time"
)
const network = "unix"
// withPeerFD runs f on the descriptor behind conn, which must be a unix socket, and
// returns whichever error came first: not a unix socket, no raw access to it, or f's
// own. It is the scaffolding every platform's peerUID shares; only the getsockopt inside
// differs.
func withPeerFD(conn net.Conn, f func(fd int) error) error {
uc, ok := conn.(*net.UnixConn)
if !ok {
return fmt.Errorf("not a unix connection")
}
raw, err := uc.SyscallConn()
if err != nil {
return err
}
var inner error
if err := raw.Control(func(fd uintptr) { inner = f(int(fd)) }); err != nil {
return err
}
return inner
}
// currentUID is the user the owner checks are made for. A variable so a test can stand
// in for root, or for a stranger, without owning a second account on the host.
var currentUID = func() uint32 { return uint32(os.Geteuid()) }
// listen binds the socket at path with the given access.
//
// Access control is the socket's own mode and group, which the node is created no
// looser than and which are set after bind, together with the peer check the server
// applies on accept unless a group shares the socket, in which case the kernel's check
// is the whole of it: connect(2) needs write permission on the socket file on every
// unix we build for.
//
// The directory is not part of that, and is never chmoded: a directory a service
// manager made for us is served from at whatever mode it came with (systemd's
// RuntimeDirectoryMode= is 0755 by default), and one an operator pointed us into keeps
// whatever else it holds readable. The one we make ourselves is 0755, searchable by
// everyone, so that a nebula ctl run as another user reaches the socket and gets the
// kernel's answer, which for a member of the configured group is a connection and for
// anyone else is permission denied.
func listen(path string, acc Access) (net.Listener, error) {
// The kernel silently truncates an over long sun_path and bind then fails with
// "invalid argument", which tells an operator nothing. Say what is actually wrong.
if pathTooLong(path) {
return nil, fmt.Errorf("socket path is %d bytes, over the %d byte limit for a unix socket: %s", len(path), maxSocketPath, path)
}
dir := filepath.Dir(path)
// One level only. Every directory defaultPathFor offers sits directly under one that
// exists, and an operator who names a deeper path creates it; a missing parent is a
// mistake to report, not a tree to build. A directory made here stays when the
// socket goes: an empty one costs nothing, and the next start wants it again.
//
// The mode comes from the umask around the call, never from a chmod afterwards: a
// chmod by path could land on whatever somebody who can write the parent swapped in
// between the two calls, and a chmod through a descriptor on a root-owned directory
// they moved there, since root owns both. Born 0755 there is nothing to fix up.
// The umask is process wide and the window is one syscall; see umaskMu.
umaskMu.Lock()
old := syscall.Umask(0022)
err := os.Mkdir(dir, 0755)
syscall.Umask(old)
umaskMu.Unlock()
if err != nil && !errors.Is(err, os.ErrExist) {
return nil, fmt.Errorf("could not create %s: %w", dir, err)
}
// Whether it was made here or came with the environment, the directory has to be
// one nobody else could change; see checkSocketDir.
if err := checkSocketDir(dir); err != nil {
return nil, err
}
if err := clearStale(path); err != nil {
return nil, err
}
// bind creates the socket node at the process umask, and a connect(2) between that
// and a later chmod would be queued for accept. Tighten the umask around the bind so
// the node is never looser than 0600, however briefly; see umaskMu.
umaskMu.Lock()
old = syscall.Umask(0077)
ln, err := net.Listen(network, path)
syscall.Umask(old)
umaskMu.Unlock()
if err != nil {
return nil, err
}
// The node at path is given up by release, which unlinks it before the listener
// closes; see there for why the order matters. Go's listener would unlink on Close
// as well, and an unlink after the close could only be of whatever bound the path
// next, so that is turned off and release is the one place the node is removed.
if ul, ok := ln.(*net.UnixListener); ok {
ul.SetUnlinkOnClose(false)
}
// The umask only promises no looser than 0600; this pins the mode and the group
// exactly, and is the one place a failure to secure the socket is reported.
if err := applyAccess(path, acc); err != nil {
release(ln, path)
return nil, fmt.Errorf("could not secure %s: %w", path, err)
}
return ln, nil
}
// applyAccess sets the owning group and the mode of the socket node at path. With no
// group configured the socket is handed back to the daemon's own group, so taking the
// key out of the config on a reload takes the access away as surely as a restart would.
// The chown is Lchown, so a link at the path is not followed to its target.
func applyAccess(path string, acc Access) error {
gid, err := gidFor(acc)
if err != nil {
return err
}
if err := os.Lchown(path, -1, gid); err != nil {
return err
}
return os.Chmod(path, acc.Mode.Perm())
}
// checkSocketDir refuses to serve from a directory that somebody other than the user
// nebula runs as, or root, could change. The socket's mode and group say who may connect
// to it; what makes the path they connect to the daemon's is the directory, since
// whoever can write it can put a socket of their own at the path and answer the clients
// in the daemon's place. So the directory must be owned by the daemon's user or by root
// and writable by its owner alone; a sticky bit is not an exception. A runtime directory
// a service manager made for another user, or /var/run/nebula created ahead of a root
// daemon by a member of group daemon, where darwin lets that group write /var/run, are
// both refused with the owner named.
//
// A link where the directory should be is judged twice: whoever owns the link decides
// where the path leads, so the link must be ours or root's too, and the directory it
// resolves to is then held to the rule above. /var/run is such a link on linux, root's,
// to /run, and a socket named directly under it is served; a link somebody else planted
// is refused as theirs.
//
// Only this directory is checked, not the ones above it. /var/run itself is group
// writable on darwin by design, and the parents are the operator's to trust, as they are
// for every file nebula reads. The check is made at bind, so a directory loosened while
// the daemon runs is noticed at the next reload.
func checkSocketDir(dir string) error {
fi, err := os.Lstat(dir)
if err != nil {
return err
}
if fi.Mode()&os.ModeSymlink != 0 {
if uid, ok := owner(fi); ok && uid != currentUID() && uid != 0 {
return fmt.Errorf("%s is a symlink owned by %s, not by the user nebula runs as or by root, and its owner decides where the path leads; name a socket path that does not go through it", dir, describeUID(uid))
}
if fi, err = os.Stat(dir); err != nil {
return err
}
}
if !fi.IsDir() {
return fmt.Errorf("%s is not a directory", dir)
}
if uid, ok := owner(fi); ok && uid != currentUID() && uid != 0 {
return fmt.Errorf("%s is owned by %s, not by the user nebula runs as or by root, and its owner could replace the socket; make it ours or root's, or name a socket path in a directory that is", dir, describeUID(uid))
}
if perm := fi.Mode().Perm(); perm&0022 != 0 {
who := "its group"
if perm&0002 != 0 {
who = "everyone"
}
return fmt.Errorf("%s is writable by %s (mode %04o), who could replace the socket; make it writable by its owner alone, or name a socket path in a directory that is", dir, who, perm)
}
return nil
}
// owner returns the uid that owns the file fi describes.
func owner(fi os.FileInfo) (uint32, bool) {
st, ok := fi.Sys().(*syscall.Stat_t)
if !ok {
return 0, false
}
return uint32(st.Uid), true
}
// describeUID renders a uid with its account name when the host can say what it is, so
// a message about somebody else's socket tells the operator who that is.
func describeUID(uid uint32) string {
id := strconv.FormatUint(uint64(uid), 10)
if u, err := user.LookupId(id); err == nil && u.Username != "" {
return "uid " + id + " (" + u.Username + ")"
}
return "uid " + id
}
// clearStale removes a leftover socket file, but only after proving nothing is
// listening on it. An unclean restart leaves one behind every time, so this is a
// routine path; blindly unlinking would let a second daemon steal a live endpoint.
//
// Liveness is a plain connect. Only a refused connect proves there is nothing behind
// the file. Anything else, a timeout or a permission error, is a socket we cannot vouch
// for, and it is left alone with the reason in the error.
func clearStale(path string) error {
fi, err := os.Stat(path)
if err != nil {
if os.IsNotExist(err) {
return nil
}
return err
}
if !isSocket(fi) {
return fmt.Errorf("%s exists and is not a socket", path)
}
conn, err := dial(path, 250*time.Millisecond)
if err == nil {
_ = conn.Close()
return fmt.Errorf("%s is already in use by another nebula instance", path)
}
if errors.Is(err, os.ErrNotExist) {
// Gone between the stat and the connect; nothing to clear.
return nil
}
if !errors.Is(err, syscall.ECONNREFUSED) {
return fmt.Errorf("%s exists and could not be shown to be stale, refusing to remove it: %w", path, err)
}
return os.Remove(path)
}
// removeSocket unlinks the socket node. Its directory stays, whoever made it; see
// listen.
func removeSocket(path string) {
_ = os.Remove(path)
}
func dial(path string, timeout time.Duration) (net.Conn, error) {
return net.DialTimeout(network, path, timeout)
}
// sunPathMax is the smallest sun_path any platform we build for offers (104 on darwin,
// 108 on linux). Overrunning it truncates the path rather than failing cleanly, so we
// check against the tighter bound everywhere.
const sunPathMax = 104
// maxSocketPath is the longest path that actually binds. sun_path holds the trailing
// NUL as well, and Go's syscall package refuses a path that fills the whole buffer
// with a bare EINVAL, so the usable limit is one byte short of the buffer.
const maxSocketPath = sunPathMax - 1
// umaskMu serialises the two places the process umask is set around one syscall: the
// umask is process wide, and two endpoints binding at once, a reload landing during
// startup, would otherwise interleave their save and restore and leave the daemon's
// umask at one of ours. It is held for the syscall alone. Another goroutine creating a
// file in that moment gets the umask of the window, 0022 or 0077, rather than the
// process's own; nothing creates files on a hot path.
var umaskMu sync.Mutex
func pathTooLong(path string) bool { return len(path) > maxSocketPath }
// DefaultSocketPath is the control socket when the config names none. It is root's to
// create, as the tun device is root's to open; a daemon run as anyone else names a path
// of its own with CtlSocketKey.
const DefaultSocketPath = "/var/run/" + DirName + "/ctl.sock"
// checkSocketPath refuses a configured socket path that is relative, or too long to
// bind. A relative path would mean a different socket to the daemon and to nebula ctl,
// each resolving it against the directory it happened to be started from.
func checkSocketPath(key, value string) error {
if !filepath.IsAbs(value) {
return fmt.Errorf("%s must be an absolute path, got %q", key, value)
}
if pathTooLong(value) {
return fmt.Errorf("%s is %d bytes, too long for this platform: %s", key, len(value), value)
}
return nil
}
func isSocket(fi os.FileInfo) bool { return fi.Mode()&os.ModeSocket != 0 }
-358
View File
@@ -1,358 +0,0 @@
//go:build !windows
package localapi
import (
"bufio"
"context"
"net"
"os"
"path/filepath"
"strings"
"syscall"
"testing"
"time"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// privateBase returns a short directory of our own, 0700, for tests to lay sockets out
// under, short enough that a path under it fits sun_path.
func privateBase(t *testing.T) string {
t.Helper()
return test.ShortTempDir(t)
}
// openDir returns a world writable directory under base, the shape of /tmp.
func openDir(t *testing.T, base string) string {
t.Helper()
open := filepath.Join(base, "open")
require.NoError(t, os.Mkdir(open, 0777))
require.NoError(t, os.Chmod(open, 0777|os.ModeSticky))
return open
}
// TestListenLeavesAServiceManagersDirectoryAlone covers the directory a service manager
// pre-creates: systemd's RuntimeDirectory is 0755 by default. The socket is 0600, which
// is what refuses other users, so the directory is served from as it came and nothing
// is chmoded that the operator did not set.
func TestListenLeavesAServiceManagersDirectoryAlone(t *testing.T) {
base := privateBase(t)
dir := filepath.Join(base, DirName)
require.NoError(t, os.Mkdir(dir, 0755))
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(context.Background(), filepath.Join(dir, "s.sock"), ownerOnly()))
t.Cleanup(s.Stop)
fi, err := os.Stat(dir)
require.NoError(t, err)
assert.Equal(t, os.FileMode(0755), fi.Mode().Perm(), "the directory is left exactly as it was")
fi, err = os.Stat(s.Path())
require.NoError(t, err)
assert.Equal(t, os.FileMode(0600), fi.Mode().Perm(), "the socket is what is private")
_, err = (&Client{Path: s.Path()}).Describe()
require.NoError(t, err)
}
// TestListenCreatesOneLevelOnly: the socket directory is created when it is missing,
// but nothing above it is. A runtime directory that does not exist is the operator's to
// provide, and the failure says which directory could not be created.
func TestListenCreatesOneLevelOnly(t *testing.T) {
base := privateBase(t)
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(context.Background(), filepath.Join(base, DirName, "s.sock"), ownerOnly()))
s.Stop()
dir := filepath.Join(base, "missing", DirName)
err := s.Rebind(context.Background(), filepath.Join(dir, "s.sock"), ownerOnly())
require.Error(t, err)
require.ErrorContains(t, err, "could not create "+dir)
assert.NoDirExists(t, filepath.Join(base, "missing"))
}
// TestListenRestoresTheUmask: listen tightens the process umask around bind so the socket
// node is never created looser than 0600. That is process wide state, so it must come
// back exactly as it was, on success and on failure alike.
func TestListenRestoresTheUmask(t *testing.T) {
old := syscall.Umask(0022)
t.Cleanup(func() { syscall.Umask(old) })
base := privateBase(t)
ln, err := listen(filepath.Join(base, DirName, "s.sock"), ownerOnly())
require.NoError(t, err)
t.Cleanup(func() { ln.Close() })
assert.Equal(t, 0022, syscall.Umask(0022), "umask after a successful listen")
_, err = listen(filepath.Join(base, DirName, "s.sock"), ownerOnly())
require.Error(t, err, "the path is already bound")
assert.Equal(t, 0022, syscall.Umask(0022), "umask after a failed listen")
}
// TestListenServesFromASharedDirectory: an operator who points local_api.control.socket
// into a directory holding other files, /etc/nebula say, gets a socket there and keeps
// the config and certs readable by everyone who could read them before.
func TestListenServesFromASharedDirectory(t *testing.T) {
base := privateBase(t)
dir := filepath.Join(base, "etc-nebula")
require.NoError(t, os.Mkdir(dir, 0755))
cfg := filepath.Join(dir, "config.yml")
require.NoError(t, os.WriteFile(cfg, []byte("{}\n"), 0644))
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(context.Background(), filepath.Join(dir, "ctl.sock"), ownerOnly()))
t.Cleanup(s.Stop)
fi, err := os.Stat(dir)
require.NoError(t, err)
assert.Equal(t, os.FileMode(0755), fi.Mode().Perm())
fi, err = os.Stat(cfg)
require.NoError(t, err)
assert.Equal(t, os.FileMode(0644), fi.Mode().Perm())
}
func TestOverlongPathIsDiagnosed(t *testing.T) {
long := filepath.Join(os.TempDir(), string(make([]byte, 0)))
for len(long) <= sunPathMax {
long = filepath.Join(long, "abcdefghij")
}
s := NewServer(test.NewLogger(), basicRegistry())
err := s.Rebind(context.Background(), filepath.Join(long, "s.sock"), ownerOnly())
require.Error(t, err)
assert.ErrorContains(t, err, "byte limit for a unix socket")
}
// TestPathLimitIsExact pins the boundary: sun_path has room for the NUL as well, and Go
// refuses a path that fills the whole buffer, so the last usable length is one short.
func TestPathLimitIsExact(t *testing.T) {
assert.False(t, pathTooLong(strings.Repeat("a", maxSocketPath)))
assert.True(t, pathTooLong(strings.Repeat("a", maxSocketPath+1)))
assert.True(t, pathTooLong(strings.Repeat("a", sunPathMax)))
}
// TestNamedPathIsUsedAsGiven: a path the operator named is served from, and dialed,
// wherever it is, a directory other users can write to included; the socket's own mode
// is what keeps them out.
func TestNamedPathIsUsedAsGiven(t *testing.T) {
base := privateBase(t)
path := filepath.Join(openDir(t, base), DirName, "s.sock")
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
t.Cleanup(s.Stop)
_, err := (&Client{Path: path}).Describe()
require.NoError(t, err)
}
// TestStaleSocketIsCleared: a dead socket in a good directory is the routine leftover of
// a hard kill, and the daemon's stale check must see it as free to replace.
func TestStaleSocketIsCleared(t *testing.T) {
path := testSocket(t)
require.NoError(t, os.Mkdir(filepath.Dir(path), 0700))
ln, err := net.Listen(network, path)
require.NoError(t, err)
require.NoError(t, os.Chmod(path, 0600))
// Close the listener but keep the file, the way a hard kill does.
ln.(*net.UnixListener).SetUnlinkOnClose(false)
require.NoError(t, ln.Close())
_, statErr := os.Stat(path)
require.NoError(t, statErr, "the socket file is still there")
require.NoError(t, clearStale(path), "a stale socket is cleared, not reported as in use")
_, statErr = os.Stat(path)
assert.True(t, os.IsNotExist(statErr))
}
// TestClientSaysWhenTheReplyCannotBeRead: something answered on the socket, but not
// with a response. That is reported as a reply that could not be read, which is what
// it is, and not as a failure to reach the daemon, whether the other side read the
// request first or hung up on it: a daemon that is not nebula may do either.
func TestClientSaysWhenTheReplyCannotBeRead(t *testing.T) {
for _, readsFirst := range []bool{true, false} {
path := testSocket(t)
require.NoError(t, os.Mkdir(filepath.Dir(path), 0700))
ln, err := net.Listen(network, path)
require.NoError(t, err)
require.NoError(t, os.Chmod(path, 0600))
go func() {
c, err := ln.Accept()
if err != nil {
return
}
if readsFirst {
_, _ = bufio.NewReader(c).ReadBytes('\n')
}
_, _ = c.Write([]byte("this is not a response\n"))
c.Close()
}()
_, err = (&Client{Path: path}).Describe()
var unread *ReplyError
require.ErrorAs(t, err, &unread, "reads first: %v", readsFirst)
assert.Contains(t, err.Error(), "the daemon's reply could not be read")
ln.Close()
}
}
// TestClientGivesUpOnASilentDaemon: a daemon that accepts and then never answers must
// not hang the client for good. A script wrapping nebula ctl has no other way out.
func TestClientGivesUpOnASilentDaemon(t *testing.T) {
path := testSocket(t)
require.NoError(t, os.Mkdir(filepath.Dir(path), 0700))
ln, err := net.Listen(network, path)
require.NoError(t, err)
defer ln.Close()
require.NoError(t, os.Chmod(path, 0600))
accepted := make(chan net.Conn, 1)
go func() {
c, err := ln.Accept()
if err == nil {
accepted <- c // held open, never written to
}
}()
start := time.Now()
_, err = (&Client{Path: path, Timeout: 200 * time.Millisecond}).Describe()
require.Error(t, err)
require.ErrorContains(t, err, "did not answer within")
assert.Less(t, time.Since(start), 5*time.Second)
select {
case c := <-accepted:
_ = c.Close()
default:
}
}
// asUID makes the owner checks run as uid for the rest of the test. The socket the test
// serves is still owned by the real user, so from uid's point of view it belongs to
// somebody else: the shape of a service account's daemon, and of an impostor's.
func asUID(t *testing.T, uid uint32) {
t.Helper()
real := currentUID
currentUID = func() uint32 { return uid }
t.Cleanup(func() { currentUID = real })
}
// TestPeerCheckRefusesAnotherUser: a server built as one user and serving as another
// refuses the real user at the peer check, and the refusal reaches the client as the
// daemon's answer rather than as a failure to reach it.
func TestPeerCheckRefusesAnotherUser(t *testing.T) {
if !peerCheckSupported {
t.Skip("no peer credentials on this platform")
}
me := currentUID()
asUID(t, me+1)
s := NewServer(test.NewLogger(), basicRegistry())
asUID(t, me)
path := testSocket(t)
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
defer s.Stop()
_, err := (&Client{Path: path}).Describe()
require.Error(t, err)
require.ErrorContains(t, err, "not permitted")
var refused *DaemonError
assert.ErrorAs(t, err, &refused, "the daemon answered; this is not a reachability failure")
}
// TestListenerDoesNotUnlinkOnClose: the server unlinks its socket node itself, before
// the listener closes, and the listener must not unlink again afterwards. By the time a
// close happens the path may belong to a newer daemon that found the node stale and
// bound its own, and a second unlink would take that daemon's socket from under it.
func TestListenerDoesNotUnlinkOnClose(t *testing.T) {
path := filepath.Join(privateBase(t), DirName, "s.sock")
old, err := listen(path, ownerOnly())
require.NoError(t, err)
// A newer daemon's start, with the old node gone and the path bound afresh.
require.NoError(t, os.Remove(path))
next, err := net.Listen(network, path)
require.NoError(t, err)
t.Cleanup(func() { next.Close() })
require.NoError(t, old.Close())
_, err = os.Stat(path)
assert.NoError(t, err, "closing the old listener left the newer daemon's socket in place")
}
// TestListenRefusesADirectoryOthersCanWrite: whoever can write the socket's directory
// can put a socket of their own at the path and answer the clients, who trust the path;
// so a directory writable by its group or by everyone is refused, and nothing is bound.
func TestListenRefusesADirectoryOthersCanWrite(t *testing.T) {
for _, mode := range []os.FileMode{0775, 0757, 0777, 01777} {
dir := filepath.Join(privateBase(t), "loose")
require.NoError(t, os.Mkdir(dir, mode))
require.NoError(t, os.Chmod(dir, mode))
_, err := listen(filepath.Join(dir, "s.sock"), ownerOnly())
require.ErrorContains(t, err, "writable by", "mode %04o", mode)
assert.NoFileExists(t, filepath.Join(dir, "s.sock"))
}
}
// TestListenRefusesADirectoryOwnedBySomebodyElse: a directory that is neither ours nor
// root's is refused, since its owner could replace the socket. The test's own directory
// stands in for somebody else's by the process pretending to be another user.
func TestListenRefusesADirectoryOwnedBySomebodyElse(t *testing.T) {
me := currentUID()
if me == 0 {
t.Skip("root's directories are accepted whoever is asking")
}
dir := filepath.Join(privateBase(t), "theirs")
require.NoError(t, os.Mkdir(dir, 0755))
asUID(t, me+1)
_, err := listen(filepath.Join(dir, "s.sock"), ownerOnly())
require.ErrorContains(t, err, "is owned by")
assert.NoFileExists(t, filepath.Join(dir, "s.sock"))
asUID(t, me)
ln, err := listen(filepath.Join(dir, "s.sock"), ownerOnly())
require.NoError(t, err, "the same directory, seen as our own, serves")
release(ln, filepath.Join(dir, "s.sock"))
}
// TestListenThroughASymlinkedDirectory: a link where the directory should be is judged
// by who owns it, since they decide where the path leads, and then by the directory it
// resolves to. Our own link to a proper directory serves, as /var/run on linux is root's
// link to /run; somebody else's link is refused, and so is a link of ours to a directory
// others can write.
func TestListenThroughASymlinkedDirectory(t *testing.T) {
me := currentUID()
if me == 0 {
t.Skip("root's links are accepted whoever is asking")
}
base := privateBase(t)
real := filepath.Join(base, "real")
require.NoError(t, os.Mkdir(real, 0755))
link := filepath.Join(base, "link")
require.NoError(t, os.Symlink(real, link))
path := filepath.Join(link, "s.sock")
ln, err := listen(path, ownerOnly())
require.NoError(t, err, "our link to our directory")
release(ln, path)
assert.NoFileExists(t, filepath.Join(real, "s.sock"))
asUID(t, me+1)
_, err = listen(path, ownerOnly())
require.ErrorContains(t, err, "is a symlink owned by")
asUID(t, me)
require.NoError(t, os.Chmod(real, 0777))
_, err = listen(path, ownerOnly())
require.ErrorContains(t, err, "writable by everyone")
}
-255
View File
@@ -1,255 +0,0 @@
//go:build windows
package localapi
import (
"context"
"errors"
"fmt"
"net"
"strings"
"time"
"github.com/Microsoft/go-winio"
"golang.org/x/sys/windows"
)
// pipePrefix is where windows named pipes live. The pipe's own security descriptor
// carries the access control that the socket's mode carries on unix.
const pipePrefix = `\\.\pipe\`
// protectedPrefix is the pipe namespace that only administrators and SYSTEM may create
// names under; the kernel refuses everyone else. The default name lives there, so the
// name itself says who could have created the pipe answering it: no ordinary user can
// have put anything under it ahead of the daemon. Tailscale, WireGuard and NetBird keep
// their daemon pipes there for the same reason. nebula on windows is already an
// administrator, since creating the tun adapter requires it.
const protectedPrefix = pipePrefix + `ProtectedPrefix\Administrators\`
const network = "pipe"
// currentUID is the identity the unix owner checks are made for. There is no uid here
// and no per-connection check that would read one, so the value is never consulted; it
// exists so a Server is built the same way on every platform.
var currentUID = func() uint32 { return 0 }
// listen creates the named pipe at path with a security descriptor that admits the user
// nebula runs as, SYSTEM and Administrators, and the group acc names, if any, for
// reading and writing, which is what connecting to a pipe takes. acc's mode means
// nothing to a pipe.
//
// The first instance is created, never opened, so a name some other process already
// holds is refused rather than joined: that process would otherwise answer clients in
// nebula's place. Under protectedPrefix only an administrator could hold it; anywhere
// else any account could. Either way the refusal is reported as errEndpointHeld for the
// caller to log as the event it is, except when the refusal is the namespace's own: a
// daemon that is not an administrator cannot create under protectedPrefix at all, and
// that is told apart so the operator is sent to elevate or to name a pipe elsewhere
// rather than to look for a process that is not there.
func listen(path string, acc Access) (net.Listener, error) {
sddl, err := sddlFor(acc)
if err != nil {
return nil, fmt.Errorf("could not build a security descriptor for %s: %w", path, err)
}
ln, err := winio.ListenPipe(path, &winio.PipeConfig{
SecurityDescriptor: sddl,
InputBufferSize: int32(MaxRequestBytes),
OutputBufferSize: 64 * 1024,
})
if err != nil {
return nil, listenRefused(path, isAdministrator(), err)
}
return ln, nil
}
// listenRefused turns a failure to create the pipe at path into the error the caller
// logs. admin says whether this process may create under protectedPrefix.
//
// ERROR_ACCESS_DENIED is ambiguous: it is what the kernel says to a non-administrator
// creating under protectedPrefix, and also what an existing instance whose DACL does
// not admit us says to a second create. A process that is an administrator cannot be
// getting the first, so for it, and for every name outside the prefix, access denied
// means held.
func listenRefused(path string, admin bool, err error) error {
if errors.Is(err, windows.ERROR_ACCESS_DENIED) && isProtectedPath(path) && !admin {
return fmt.Errorf("creating %s requires an administrator, as only administrators may create pipes under %s; run nebula elevated, or set %s to a pipe name outside it: %w", path, protectedPrefix, CtlSocketKey, err)
}
if errors.Is(err, windows.ERROR_ALREADY_EXISTS) || errors.Is(err, windows.ERROR_ACCESS_DENIED) {
return fmt.Errorf("%w: %s: %w", errEndpointHeld, path, err)
}
return err
}
// isProtectedPath says whether path is a name under protectedPrefix. Pipe names are
// case insensitive, so the comparison is too.
func isProtectedPath(path string) bool {
return len(path) > len(protectedPrefix) && strings.EqualFold(path[:len(protectedPrefix)], protectedPrefix)
}
// isAdministrator says whether this process holds Administrators as an enabled group,
// which is what creating under protectedPrefix takes. SYSTEM holds it; an elevated
// administrator holds it; an administrator's unelevated token carries it deny-only,
// which the membership check correctly reads as no. The membership is checked against
// the thread's own token: CheckTokenMembership wants an impersonation token and makes
// one from the primary token when handed none. A check that cannot be made is read as
// not an administrator, which only changes the wording of a failure that has already
// happened.
func isAdministrator() bool {
admins, err := windows.CreateWellKnownSid(windows.WinBuiltinAdministratorsSid)
if err != nil {
return false
}
member, err := windows.Token(0).IsMember(admins)
return err == nil && member
}
// sddlFor grants full access to the account nebula is running as, to SYSTEM and to
// Administrators, and pipeClientAccess to the group acc names, if any; nobody else. P
// makes the DACL protected, so it does not inherit entries that would widen it. It is
// dockerd's descriptor for its pipe, with its -G group in the same place, except for
// the group's mask; see pipeClientAccess.
//
// Administrators is the counterpart of the uid 0 allowance in the unix peer check. The
// service normally runs as LocalSystem, so without it an operator in an elevated shell
// could not reach the daemon at all; and an administrator can already stop the service,
// so this grants no authority they did not have.
func sddlFor(acc Access) (string, error) {
token := windows.GetCurrentProcessToken()
user, err := token.GetTokenUser()
if err != nil {
return "", err
}
sddl := fmt.Sprintf("D:P(A;;GA;;;%s)(A;;GA;;;SY)(A;;GA;;;BA)", user.User.Sid.String())
if acc.shared() {
sddl += fmt.Sprintf("(A;;0x%x;;;%s)", pipeClientAccess, acc.Group.ID)
}
return sddl, nil
}
// pipeClientAccess is what a client needs on the pipe to talk to the daemon, and all a
// configured group is granted: read and write the data, read the pipe's attributes
// and its security descriptor, which the owner check needs, and wait on it. Not the
// generic rights. On a named pipe GENERIC_WRITE carries FILE_CREATE_PIPE_INSTANCE, the
// right to put a second instance under the same name, and a group member with that
// could sit on the name and answer other clients in the daemon's place; the client's
// owner check would not catch it, since the instance it reached would be the
// member's. A client therefore has to ask for exactly this, as dial does, and a
// program that opens the info pipe with generic read and write is refused.
const pipeClientAccess uint32 = windows.FILE_READ_DATA | windows.FILE_WRITE_DATA |
windows.FILE_READ_ATTRIBUTES | windows.READ_CONTROL | windows.SYNCHRONIZE
// applyAccess cannot change a pipe's access in place: its security descriptor is fixed
// when it is created, so a change means creating it again, which the caller does.
func applyAccess(path string, acc Access) error { return errReopenForAccess }
// removeSocket is a no-op: a named pipe disappears when its last handle closes, so
// there is nothing to unlink and no stale instance to clean up.
func removeSocket(path string) {}
// dial connects to the pipe and, before the connection carries anything, checks who
// serves it. The request names a command and its arguments, so it must not reach a
// pipe that some other account put under nebula's name. For the default name the
// namespace already promises that, since only an administrator could have created it;
// the check is what covers a CtlSocketKey override outside protectedPrefix, and costs
// nothing to make on every name.
func dial(path string, timeout time.Duration) (net.Conn, error) {
ctx, cancel := context.WithTimeout(context.Background(), timeout)
defer cancel()
conn, err := winio.DialPipeAccess(ctx, path, pipeClientAccess)
if err != nil {
return nil, err
}
if err := checkPipeOwner(conn, path); err != nil {
conn.Close()
return nil, err
}
return conn, nil
}
// checkPipeOwner refuses a pipe whose owner is not an account that could be running
// nebula. It is the unix owner check made from the other side: the handle winio dialed
// exposes the pipe object, and the object's owner is whoever created it.
//
// The owner is what makes this sound. An account can only create an object owned by
// itself, or by a SID it holds with the owner flag, unless it has SeRestorePrivilege,
// which ordinary users do not; so a pipe owned by SYSTEM, Administrators or a service
// account was put there by one of them, and never by a user squatting the name. winio
// dials with SECURITY_ANONYMOUS, so the impostor could not have impersonated the client
// either, but it could have read the command and answered it with whatever it liked.
func checkPipeOwner(conn net.Conn, path string) error {
h, ok := conn.(interface{ Fd() uintptr })
if !ok {
return fmt.Errorf("%s: cannot identify who serves the pipe", path)
}
sd, err := windows.GetSecurityInfo(windows.Handle(h.Fd()), windows.SE_KERNEL_OBJECT, windows.OWNER_SECURITY_INFORMATION)
if err != nil {
return fmt.Errorf("%s: could not read the pipe's owner: %w", path, err)
}
owner, _, err := sd.Owner()
if err != nil {
return fmt.Errorf("%s: could not read the pipe's owner: %w", path, err)
}
if owner == nil {
return fmt.Errorf("%s has no owner; refusing to send to it", path)
}
me, err := windows.GetCurrentProcessToken().GetTokenUser()
if err != nil {
return fmt.Errorf("%s: could not read own identity: %w", path, err)
}
if ownerAccepted(owner, me.User.Sid) {
return nil
}
return fmt.Errorf("%s is served by %s, not by nebula; run nebula ctl as that user, or pass -config or -socket to name the daemon's pipe", path, describeSID(owner))
}
// ownerAccepted says whether a pipe owned by owner may be sent a request by a client
// running as me: the client's own account, and the two accounts a daemon runs under
// that no ordinary user can create an object as, SYSTEM and Administrators. Another
// account is refused, whoever the client is; the daemon's DACL decides who may connect,
// and this decides who may answer.
func ownerAccepted(owner, me *windows.SID) bool {
return owner.Equals(me) ||
owner.IsWellKnown(windows.WinLocalSystemSid) ||
owner.IsWellKnown(windows.WinBuiltinAdministratorsSid)
}
// describeSID renders a SID with its account name when the host can say what it is, so
// a message about somebody else's pipe tells the operator who that is.
func describeSID(sid *windows.SID) string {
account, domain, _, err := sid.LookupAccount("")
if err != nil || account == "" {
return sid.String()
}
if domain == "" {
return account
}
return domain + `\` + account
}
// DefaultSocketPath is the control pipe when the config names none. It is under
// protectedPrefix, so whatever answers it was created by an administrator or SYSTEM; see
// there.
const DefaultSocketPath = protectedPrefix + "nebula-ctl"
// checkSocketPath refuses a configured name that is not a full pipe name. A pipe lives
// under pipePrefix and nowhere else, so a bare name has no meaning here, and one under
// pipePrefix is used as given. An override outside protectedPrefix is a name any
// account could have created first; the owner check in dial is what protects a client
// of one.
func checkSocketPath(key, value string) error {
if len(value) <= len(pipePrefix) || !strings.EqualFold(value[:len(pipePrefix)], pipePrefix) {
return fmt.Errorf(`%s must be a full pipe name under %s, got %q`, key, pipePrefix, value)
}
return nil
}
-186
View File
@@ -1,186 +0,0 @@
//go:build windows
package localapi
import (
"errors"
"net"
"testing"
"time"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"golang.org/x/sys/windows"
)
// TestOwnerAccepted pins which accounts may answer a nebula ctl request: the client's
// own, and the accounts a daemon runs as that an ordinary user cannot create an object
// owned by. Any other user is refused, since any user can create any pipe name.
func TestOwnerAccepted(t *testing.T) {
me, err := windows.GetCurrentProcessToken().GetTokenUser()
require.NoError(t, err)
wellKnown := func(typ windows.WELL_KNOWN_SID_TYPE) *windows.SID {
sid, err := windows.CreateWellKnownSid(typ)
require.NoError(t, err)
return sid
}
parse := func(s string) *windows.SID {
sid, err := windows.StringToSid(s)
require.NoError(t, err)
return sid
}
accepted := map[string]*windows.SID{
"self": me.User.Sid,
"SYSTEM": wellKnown(windows.WinLocalSystemSid),
"Administrators": wellKnown(windows.WinBuiltinAdministratorsSid),
}
for name, sid := range accepted {
assert.True(t, ownerAccepted(sid, me.User.Sid), name)
}
refused := map[string]*windows.SID{
"Everyone": wellKnown(windows.WinWorldSid),
"Users": wellKnown(windows.WinBuiltinUsersSid),
"another user": parse("S-1-5-21-1-2-3-1001"),
"a service's own account": parse("S-1-5-80-1-2-3-4-5"),
"LocalService": wellKnown(windows.WinLocalServiceSid),
}
for name, sid := range refused {
assert.False(t, ownerAccepted(sid, me.User.Sid), name)
}
}
// TestDefaultPathIsProtected pins the default name to the namespace only administrators
// and SYSTEM can create under, which is what lets nebula ctl trust the default name
// from its spelling alone.
func TestDefaultPathIsProtected(t *testing.T) {
p := DefaultSocketPath
assert.True(t, isProtectedPath(p), p)
assert.Equal(t, `\\.\pipe\ProtectedPrefix\Administrators\nebula-ctl`, p)
}
// TestConfiguredPipeNameMustBeFull: a configured name is used as given, and only a full
// pipe name means anything here.
func TestConfiguredPipeNameMustBeFull(t *testing.T) {
p, err := SocketPathFromConfig(fakeWinConfig{CtlSocketKey: `\\.\pipe\nebula-ctl`})
require.NoError(t, err)
assert.Equal(t, `\\.\pipe\nebula-ctl`, p)
for _, bad := range []string{"nebula-ctl", `C:\nebula\ctl.sock`, `\\.\pipe\`} {
_, err := SocketPathFromConfig(fakeWinConfig{CtlSocketKey: bad})
require.Error(t, err, bad)
assert.ErrorContains(t, err, "full pipe name", bad)
}
}
type fakeWinConfig map[string]string
func (f fakeWinConfig) GetString(k, d string) string {
if v, ok := f[k]; ok {
return v
}
return d
}
func TestIsProtectedPath(t *testing.T) {
assert.True(t, isProtectedPath(`\\.\pipe\ProtectedPrefix\Administrators\nebula-1`))
assert.True(t, isProtectedPath(`\\.\pipe\protectedprefix\administrators\nebula-1`), "pipe names are case insensitive")
assert.False(t, isProtectedPath(`\\.\pipe\ProtectedPrefix\Administrators\`), "the prefix alone names nothing")
assert.False(t, isProtectedPath(`\\.\pipe\nebula-1`))
assert.False(t, isProtectedPath(`\\.\pipe\ProtectedPrefix\Users\nebula-1`))
}
// TestListenRefused pins how a failed create is reported. Access denied under the
// protected prefix is the namespace refusing a non-administrator, and says so; the same
// error for an administrator, or for any name outside the prefix, is a name some other
// process holds, and is errEndpointHeld so the daemon logs it as an error.
func TestListenRefused(t *testing.T) {
protected := `\\.\pipe\ProtectedPrefix\Administrators\nebula-1`
plain := `\\.\pipe\nebula-1`
err := listenRefused(protected, false, windows.ERROR_ACCESS_DENIED)
assert.NotErrorIs(t, err, errEndpointHeld)
assert.ErrorIs(t, err, windows.ERROR_ACCESS_DENIED)
assert.Contains(t, err.Error(), "requires an administrator")
assert.Contains(t, err.Error(), CtlSocketKey)
assert.ErrorIs(t, listenRefused(protected, true, windows.ERROR_ACCESS_DENIED), errEndpointHeld)
assert.ErrorIs(t, listenRefused(plain, false, windows.ERROR_ACCESS_DENIED), errEndpointHeld)
assert.ErrorIs(t, listenRefused(protected, false, windows.ERROR_ALREADY_EXISTS), errEndpointHeld)
assert.ErrorIs(t, listenRefused(plain, true, windows.ERROR_ALREADY_EXISTS), errEndpointHeld)
other := errors.New("something else")
assert.Same(t, other, listenRefused(protected, false, other))
}
// TestListenProtectedName creates and dials a name under the protected prefix. It needs
// an administrator, and otherwise checks only that the refusal is the one listen
// promises a non-administrator.
func TestListenProtectedName(t *testing.T) {
path := protectedPrefix + "nebula-test-" + t.Name()
ln, err := listen(path, ownerOnly())
if !isAdministrator() {
require.Error(t, err)
assert.Contains(t, err.Error(), "requires an administrator")
t.Skip("not an administrator; the protected namespace refused the create as it should")
}
require.NoError(t, err)
t.Cleanup(func() { ln.Close() })
accepted := make(chan error, 1)
go func() {
conn, err := ln.Accept()
if err == nil {
conn.Close()
}
accepted <- err
}()
conn, err := dial(path, time.Second)
require.NoError(t, err)
conn.Close()
require.NoError(t, <-accepted)
}
// TestDialAcceptsOwnPipe dials a pipe this process created outside the protected
// prefix, as an override might name, which the owner check must pass: the pipe's owner
// is this account, or Administrators when the test is elevated.
func TestDialAcceptsOwnPipe(t *testing.T) {
path := pipePrefix + "nebula-test-" + t.Name()
ln, err := listen(path, ownerOnly())
require.NoError(t, err)
t.Cleanup(func() { ln.Close() })
accepted := make(chan error, 1)
go func() {
conn, err := ln.Accept()
if err == nil {
conn.Close()
}
accepted <- err
}()
conn, err := dial(path, time.Second)
require.NoError(t, err)
conn.Close()
require.NoError(t, <-accepted)
}
// TestListenRefusesHeldName pins that a name another listener already holds is refused
// with errEndpointHeld rather than joined, which is what keeps an impostor from being
// served alongside, or in place of, the daemon.
func TestListenRefusesHeldName(t *testing.T) {
path := pipePrefix + "nebula-test-" + t.Name()
first, err := listen(path, ownerOnly())
require.NoError(t, err)
t.Cleanup(func() { first.Close() })
var second net.Listener
second, err = listen(path, ownerOnly())
if second != nil {
second.Close()
}
require.ErrorIs(t, err, errEndpointHeld)
}
-167
View File
@@ -1,167 +0,0 @@
// Package localapi serves the nebula command registry over a local IPC endpoint so the
// user who ran nebula can inspect and control it from a CLI on the same machine.
//
// Access is the socket's own: without a group it is 0600 to the user who ran nebula,
// and the server checks the peer's credentials on every connection besides; with a
// group configured it is 0660 to that group, and the kernel's check on connect is the
// whole of it, so a member has the same control of the daemon as its user. Every
// command in the registry is served; nothing is filtered on either side.
//
// # Protocol
//
// One JSON object in, one JSON object out, both newline terminated. It is an API, not a
// command line: a request names a command and passes its options as named parameters,
// rather than shipping an argv the daemon has to parse. A client that is itself a CLI
// asks for the schema first (Request.Describe) and turns the user's flags into those
// parameters; a client that is not need never think in flags at all.
package localapi
import (
"bufio"
"bytes"
"encoding/json"
"errors"
"fmt"
"io"
)
// ProtocolVersion is the version of the request and response envelopes. A client sending
// a version this server does not know gets a sentence back rather than a parse error,
// which matters because the daemon and the CLI are upgraded independently.
const ProtocolVersion = 1
// MaxRequestBytes bounds the request. A request names a command and its parameters, so it
// has no business being large, and an unbounded read is a way for a local process to
// consume memory the packet path needs.
const MaxRequestBytes = 64 * 1024
// Request is the single JSON object a client sends.
type Request struct {
Version int `json:"version"`
// Describe asks for the schema of every command this surface serves, in place of
// running one. Command, Params and Args are ignored.
Describe bool `json:"describe,omitempty"`
// Command names the command to run.
Command string `json:"command,omitempty"`
// Params carries the command's options by name, with JSON's own types, true rather
// than "true", 3 rather than "3", or as the strings a command line parsed them from,
// which is what nebula ctl sends. Unknown names are an error, so a typo is reported
// instead of ignored. The schema lists json and pretty on every command because the
// command lines take them; here they are refused, since Format is how a request
// chooses its rendering.
Params map[string]any `json:"params,omitempty"`
// Args carries the command's operands, the values that are not options: the vpn
// address `print-tunnel` prints, for instance. They stay positional because the
// commands do not name them.
Args []string `json:"args,omitempty"`
// Format asks for a rendering other than the default. "" and "json" return the
// command's result as a JSON value, which is what a program wants. "text" asks the
// daemon to run the command's own human rendering and return it as a string, which is
// how `nebula ctl` shows an operator the same output the ssh server would.
Format string `json:"format,omitempty"`
}
// Response is the single JSON object a server sends back.
type Response struct {
Version int `json:"version"`
// Result is what a request for JSON produced: the command's value, or the schema
// when the request was a Describe, as a JSON value, so a caller reads it directly
// instead of parsing a string that happens to contain JSON.
//
// Whitespace inside a JSON result is not preserved: an envelope normalises it, and
// in an API it carries no meaning anyway. Formatting a result for a human to read is
// the client's job.
Result json.RawMessage `json:"result,omitempty"`
// Text is what a request for text produced: the command's own human rendering, the
// bytes it wrote and nothing else. They travel as bytes, base64 in the envelope,
// rather than as a JSON string, because a string is UTF-8 and the rendering need
// not be: a name from a certificate is whatever the certificate holds, and the ssh
// server passes it through, so this does too.
Text []byte `json:"text,omitempty"`
// Error explains a failure in a sentence, and is absent on success. It is the only
// signal a caller needs: there is no separate status field, because a command that
// failed always says why and one that succeeded never does. A command line client
// turns its presence into a non-zero exit.
//
// A command that ran and refused sets Error too, alongside its Result when the
// caller asked for JSON. That is how a refusal and a request that never ran are
// told apart: the first has both, the second has Error alone.
Error string `json:"error,omitempty"`
}
// frameError is a frame that arrived and was not one: oversized, or not the JSON object
// the protocol defines. It is told from a failed read so the client can say the daemon
// answered with something it could not read, rather than that it could not be reached.
type frameError struct{ err error }
func (e *frameError) Error() string { return e.err.Error() }
func (e *frameError) Unwrap() error { return e.err }
// readFrame reads one frame, a newline terminated JSON object, from r into v. With
// writeFrame it is where the framing rule lives, for both directions: a frame is a
// single line, a line over max bytes is refused rather than buffered, and a final line
// with no newline before EOF still counts. what names the frame in errors ("request", "response"); a
// read failure is returned wrapped, so a caller can still see io.EOF or a timeout in it.
// strict refuses a field v does not declare, which is the daemon's reading of a request:
// a client that misspells one deserves to hear about it, as it does for a parameter.
// The client reads a response leniently, so a field a newer daemon adds is passed over
// rather than fatal.
func readFrame(r io.Reader, max int, what string, v any, strict bool) error {
br := bufio.NewReader(io.LimitReader(r, int64(max)+1))
line, err := br.ReadBytes('\n')
if err != nil && (len(line) == 0 || !errors.Is(err, io.EOF)) {
// EOF after a partial line is the unterminated final line the rule allows, and
// an oversized line looks the same once the LimitReader cuts it off, so both go
// on to be judged by their length. Anything else is a failed read, whatever was
// buffered first: a deadline that lands mid frame is a timeout, not a parse
// error in half a document.
return fmt.Errorf("could not read %s: %w", what, err)
}
// The limit is on the body, not the frame, so the terminator does not count
// against it. A body that exceeds max is cut off by the LimitReader before its
// newline, so a missing terminator is also what an oversized line looks like.
body := bytes.TrimSuffix(line, []byte{'\n'})
if len(body) > max {
return &frameError{fmt.Errorf("%s exceeded %d bytes", what, max)}
}
dec := json.NewDecoder(bytes.NewReader(body))
if strict {
dec.DisallowUnknownFields()
}
if err := dec.Decode(v); err != nil {
return &frameError{fmt.Errorf("could not parse %s: %w", what, err)}
}
// Strict is one object and nothing after it: a second object, or stray bytes, on
// the line would be read as part of a request that happened to decode, and is
// refused the way an unknown field is.
if strict {
if _, err := dec.Token(); err != io.EOF {
return fmt.Errorf("could not parse %s: unexpected data after the object", what)
}
}
return nil
}
// writeFrame writes v to w as one frame: the JSON object and the newline that ends it.
func writeFrame(w io.Writer, v any) error {
body, err := json.Marshal(v)
if err != nil {
return fmt.Errorf("could not encode frame: %w", err)
}
_, err = w.Write(append(body, '\n'))
return err
}
-102
View File
@@ -1,102 +0,0 @@
package localapi
import (
"bytes"
"fmt"
"io"
"net"
"strings"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestReadFrameLimitIsOnTheBody pins the framing rule at its edge: the limit bounds the
// JSON body, so the newline that terminates the frame must not count against it.
func TestReadFrameLimitIsOnTheBody(t *testing.T) {
// A JSON string whose encoded form is exactly n bytes, with or without a terminator.
body := func(n int) string {
return fmt.Sprintf(`"%s"`, strings.Repeat("a", n-2))
}
var v string
// Exactly at the limit, terminated: accepted.
v = ""
require.NoError(t, readFrame(bytes.NewReader([]byte(body(64)+"\n")), 64, "request", &v, true))
assert.Len(t, v, 62)
// Exactly at the limit, EOF instead of a newline: still accepted.
v = ""
require.NoError(t, readFrame(bytes.NewReader([]byte(body(64))), 64, "request", &v, true))
assert.Len(t, v, 62)
// One byte over: refused as oversized, not as a parse failure.
err := readFrame(bytes.NewReader([]byte(body(65)+"\n")), 64, "request", &v, true)
require.Error(t, err)
assert.Contains(t, err.Error(), "exceeded 64 bytes")
}
// stallingReader hands out its bytes and then fails with err, the shape of a peer that
// sends part of a frame and then stops, until the deadline on the connection lands.
type stallingReader struct {
data []byte
err error
}
func (r *stallingReader) Read(p []byte) (int, error) {
if len(r.data) == 0 {
return 0, r.err
}
n := copy(p, r.data)
r.data = r.data[n:]
return n, nil
}
type timeoutError struct{}
func (timeoutError) Error() string { return "i/o timeout" }
func (timeoutError) Timeout() bool { return true }
func (timeoutError) Temporary() bool { return true }
// TestReadFramePartialLineThenReadErrorIsTheError: a read that fails after part of a
// frame has arrived is a failed read, and the caller sees that error. The old behaviour
// treated the fragment as the whole frame and reported a parse failure, so a client
// that stalled mid request was told its JSON was malformed, and a daemon that stalled
// mid response hid the timeout the client's error path looks for.
func TestReadFramePartialLineThenReadErrorIsTheError(t *testing.T) {
var v map[string]any
err := readFrame(&stallingReader{data: []byte(`{"command": "hel`), err: timeoutError{}}, 1024, "request", &v, true)
require.Error(t, err)
var ne net.Error
require.ErrorAs(t, err, &ne, "the timeout is still visible through the wrapping: %v", err)
assert.True(t, ne.Timeout())
assert.NotContains(t, err.Error(), "parse", "a stalled peer is not a parse failure")
// EOF after an unterminated line is the one exception: a final line with no newline
// is still a frame.
v = nil
require.NoError(t, readFrame(&stallingReader{data: []byte(`{"command": "help"}`), err: io.EOF}, 1024, "request", &v, true))
assert.Equal(t, "help", v["command"])
// And an unterminated line that is EOF only because the limit cut it off is still
// reported as oversized.
err = readFrame(bytes.NewReader([]byte(strings.Repeat("a", 70)+"\n")), 64, "request", &v, true)
require.Error(t, err)
assert.Contains(t, err.Error(), "exceeded 64 bytes")
}
// TestReadFrameUnknownFieldIsRefusedWhenStrict: the daemon refuses a request field it
// does not know, as it refuses an unknown parameter, so a misspelling is reported rather
// than ignored; the client, reading leniently, passes over a field a newer daemon sends.
func TestReadFrameUnknownFieldIsRefusedWhenStrict(t *testing.T) {
var req Request
err := readFrame(bytes.NewReader([]byte(`{"version":1,"comand":"help"}`+"\n")), 1024, "request", &req, true)
require.ErrorContains(t, err, `unknown field "comand"`)
var res Response
require.NoError(t, readFrame(bytes.NewReader([]byte(`{"version":1,"later":true}`+"\n")), 1024, "response", &res, false))
assert.Equal(t, 1, res.Version)
}
-570
View File
@@ -1,570 +0,0 @@
package localapi
import (
"bytes"
"context"
"encoding/json"
"errors"
"fmt"
"log/slog"
"math"
"net"
"runtime/debug"
"strconv"
"sync"
"sync/atomic"
"time"
"github.com/slackhq/nebula/commands"
)
const (
// defaultMaxConns caps concurrent connections. Exceeding it is rejected rather than
// queued: a queue turns a load problem into a latency problem and hides it.
defaultMaxConns = 8
// readTimeout bounds how long a client may take to send its request line.
readTimeout = 10 * time.Second
// writeTimeout bounds a single write. A client that stops reading is dropped rather
// than waited on, so a paused terminal cannot pin daemon resources.
writeTimeout = 30 * time.Second
// defaultSlowAfter is how long a command may run before the daemon says so. A
// command holds a connection slot until it returns, and one that never returns
// holds it for good; after enough of those the socket answers only "busy". The
// warning is what tells an operator that from load, and names the command.
defaultSlowAfter = 30 * time.Second
)
// Server serves a command registry over a local socket to the user who ran nebula.
type Server struct {
l *slog.Logger
src *commands.Registry
maxConns int
slowAfter time.Duration
mu sync.Mutex
ln net.Listener
path string
running bool
unwatch func() bool
sem chan struct{}
euid uint32
// access is what the socket at path was bound with, so a reload that changes only
// the group or the mode is applied without moving the socket.
access Access
// trustSocket is set while a group shares the socket. The kernel admitted the peer
// through the socket's mode and group, which is the whole access control then, and
// the peer check in serve stands aside: on unix it sees the peer's uid and primary
// gid only, so it could not tell a member of the group from anyone else.
trustSocket atomic.Bool
}
// NewServer returns a Server that will serve every command in reg, including one
// registered on it after the server is built.
func NewServer(l *slog.Logger, reg *commands.Registry) *Server {
return &Server{
l: l,
src: reg,
maxConns: defaultMaxConns,
slowAfter: defaultSlowAfter,
// The same identity the client's owner check is made for, so the peer check
// cannot disagree with it, and so a test can stand in for it the same way.
euid: currentUID(),
}
}
// Rebind brings the server up at path, moves a running server there, and serves it until
// Stop or ctx cancellation. It is the one way the server comes up, so Wire's first
// attempt and its reload callback are the same call, and a server already serving path
// is left as it is: Wire registers the callback before the first attempt, so a HUP that
// lands between the two has the callback do the first bind, and the first attempt then
// finds the socket it was about to create already up.
//
// The new socket is bound before the old one is given up: a bad path in a reloaded
// config leaves the operator with the endpoint they already had rather than none.
//
// s.mu is held for the whole operation, bind included. Releasing it around listen would
// let a Stop land in between, and this would then install a fresh listener on a server
// that had just shut down, leaving a live socket on disk until the process exits. listen
// does not take s.mu, so holding it costs the other methods one bind's wait, at most the
// stale socket probe, once per reload.
//
// An error here is never fatal to the caller: nebula keeps forwarding packets without a
// control endpoint, and the caller is expected to log and carry on.
func (s *Server) Rebind(ctx context.Context, path string, acc Access) error {
s.mu.Lock()
defer s.mu.Unlock()
// Set before the access is applied, not after, so no connection the new access
// admits is checked against the old setting. Early is safe either way: a member
// arriving before a group's mode is set is refused by the kernel, and with a group
// being taken away the peer check applies a moment before the mode tightens.
prev := s.trustSocket.Swap(acc.shared())
err := s.rebindLocked(ctx, path, acc)
if err != nil {
// Whatever was serving still is, with the access it had.
s.trustSocket.Store(prev)
}
return err
}
func (s *Server) rebindLocked(ctx context.Context, path string, acc Access) error {
reopening := false
if s.running && s.path == path {
if s.access == acc {
return nil
}
// Only the access changed. On unix it is set on the socket in place; on windows
// the descriptor is fixed when the pipe is created, so the pipe is created again.
// The name cannot be held twice, so the old one goes first, and a bind that then
// fails leaves the server down rather than on the access the operator asked to
// change; that is reported.
err := applyAccess(path, acc)
if err == nil {
s.access = acc
s.l.Info("Changed the local control socket's access", "path", path, "access", acc.String())
return nil
}
if !errors.Is(err, errReopenForAccess) {
return fmt.Errorf("could not change the access of %s: %w", path, err)
}
reopening = true
release(s.ln, s.path)
s.ln, s.path, s.running, s.access = nil, "", false, Access{}
}
ln, err := listen(path, acc)
if err != nil {
if reopening {
return fmt.Errorf("the local control socket was closed to change its access and could not be created again: %w", err)
}
return err
}
// A reload that arrives during shutdown must not bring the endpoint back. Stop is
// not terminal (Rebind after Stop is allowed, and is how a server whose first bind
// failed comes up), so running is no guide; the context is what shutdown cancels,
// and it is checked here, while nothing else can change the server's state.
if err := ctx.Err(); err != nil {
release(ln, path)
return fmt.Errorf("not starting the local control socket, nebula is shutting down: %w", err)
}
oldLn, oldPath := s.ln, s.path
s.ln = ln
s.path = path
s.access = acc
s.running = true
sem := s.semaphore()
s.watchLocked(ctx)
// Stop accepting on the old listener. In flight commands keep their connections,
// which stay usable until they finish. Its accept loop exits on the closed-listener
// error, so it is gone as soon as this returns. oldPath differs from path whenever
// oldLn is set: a running server on the same path returned above.
if oldLn != nil {
release(oldLn, oldPath)
}
s.l.Info("Local control socket listening", "path", path, "access", acc.String())
go func() {
s.accept(ln, sem)
}()
return nil
}
// watchLocked stops the server when ctx is cancelled. One watcher covers a run of the
// server, from the first bind that succeeds until Stop: a daemon whose initial listen
// failed and which later Rebinds onto a good path from a reloaded config still needs its
// socket removed on shutdown, and a Rebind on a running server must not stack a second
// watcher on the same ctx. Stop detaches the watcher, so a server started again
// afterwards is watched on the ctx that started it, and the ctx of an earlier run no
// longer reaches it whenever that one ends. Callers hold s.mu.
func (s *Server) watchLocked(ctx context.Context) {
if s.unwatch != nil {
return
}
s.unwatch = context.AfterFunc(ctx, s.Stop)
}
// Path returns the socket path currently being served, or "" when stopped.
func (s *Server) Path() string {
s.mu.Lock()
defer s.mu.Unlock()
return s.path
}
// Stop closes the listener and removes the socket. In flight commands are allowed to
// finish; Stop does not wait for them.
//
// The whole of it runs under s.mu, listener and socket included, so a Rebind cannot land
// between the state change and the close and find a path that is still bound by the
// listener it is replacing.
func (s *Server) Stop() {
s.mu.Lock()
defer s.mu.Unlock()
if !s.running {
return
}
release(s.ln, s.path)
s.running = false
s.access = Access{}
s.ln = nil
s.path = ""
// Set whenever running is. When this Stop is the watcher's own, stopping a watcher
// that has already fired is a no-op.
s.unwatch()
s.unwatch = nil
}
// release gives up a listener and the socket node at its path. The node goes first,
// while the listener is still open: for as long as it is, the node is this server's,
// since a starting daemon's stale check refuses to remove a socket that answers. The
// other order opens a window between the close and the unlink in which a new daemon can
// find the node stale, clear it and bind its own at the same path, which the unlink
// then takes from under it, leaving that daemon serving a socket nothing can reach.
// listen turns off the listener's own unlink on close so this is the only unlink.
func release(ln net.Listener, path string) {
removeSocket(path)
_ = ln.Close()
}
// semaphore returns the connection limiter, creating it on first use. It is made once
// and never replaced: a handler releases into the channel it acquired from, so swapping
// in a fresh one across a Stop and Start would leave any handler still in flight
// blocked forever on a channel nothing had filled. Callers hold s.mu.
func (s *Server) semaphore() chan struct{} {
if s.sem == nil {
s.sem = make(chan struct{}, s.maxConns)
}
return s.sem
}
// Accept failures that are not the listener closing are retried after a pause that
// doubles from acceptRetryMin up to acceptRetryMax, the same shape as net/http.Serve.
const (
acceptRetryMin = 5 * time.Millisecond
acceptRetryMax = time.Second
)
// accept serves connections from ln until it is closed. sem is the limiter to acquire
// from and release into; it is passed rather than read from s so the handlers spawned
// here never touch s.sem without the lock.
//
// Only a closed listener ends the loop; Stop and Rebind are the two things that close
// it. Every other error is retried: the process being out of descriptors (EMFILE,
// ENFILE) or a client that hung up between the kernel's queue and us (ECONNABORTED)
// are passing conditions, and giving up on them would leave a socket on disk that
// nothing answers, with the reason buried in a log line from minutes or days earlier.
func (s *Server) accept(ln net.Listener, sem chan struct{}) {
var delay time.Duration
for {
conn, err := ln.Accept()
if err != nil {
if errors.Is(err, net.ErrClosed) {
return
}
if ne, ok := err.(net.Error); ok && ne.Timeout() {
continue
}
if delay == 0 {
delay = acceptRetryMin
} else {
delay = min(delay*2, acceptRetryMax)
}
s.l.Warn("Local control socket could not accept a connection, retrying", "error", err, "retry_in", delay)
time.Sleep(delay)
continue
}
delay = 0
select {
case sem <- struct{}{}:
default:
// At capacity. Say so and hang up rather than holding the connection open.
// Logged too: a slot is held for as long as a command runs, and the slow
// command warning in serve says which ones are still running.
s.l.Warn("Local control socket is at capacity, refusing a connection", "max", s.maxConns)
s.reply(conn, &Response{Version: ProtocolVersion, Error: "local control socket is busy, try again"})
_ = conn.Close()
continue
}
go func() {
defer func() { <-sem }()
s.handle(conn)
}()
}
}
// handle serves one request. It recovers from panics in command callbacks: this
// surface is present on every install and reachable by any local process running as
// the daemon's user, so a malformed argument must cost a connection rather than the
// overlay.
func (s *Server) handle(conn net.Conn) {
defer conn.Close()
var args []string
defer func() {
if r := recover(); r != nil {
s.l.Error("Recovered from panic serving a local control command",
"panic", r,
"args", args,
"stack", string(debug.Stack()),
)
s.reply(conn, &Response{Version: ProtocolVersion, Error: "command failed"})
}
}()
// Where the platform can identify the caller, check it, unless a group shares the
// socket; see trustSocket. Otherwise listen's mode or security descriptor stands.
if peerCheckSupported && !s.trustSocket.Load() {
uid, err := peerUID(conn)
if err != nil {
s.l.Warn("Refusing local control connection with unreadable credentials", "error", err)
return
}
// Access is restricted to the user who ran nebula. Root is allowed through
// because it can reach the daemon's memory regardless of what this check says.
if uid != s.euid && uid != 0 {
s.l.Warn("Refusing local control connection from another user", "peerUid", uid)
s.reply(conn, &Response{Version: ProtocolVersion, Error: "not permitted"})
return
}
}
_ = conn.SetReadDeadline(time.Now().Add(readTimeout))
req, err := readRequest(conn)
if err != nil {
s.reply(conn, &Response{Version: ProtocolVersion, Error: err.Error()})
return
}
args = req.Args
// No deadline while a command runs; the write deadline below is what bounds a client
// that stops reading.
_ = conn.SetReadDeadline(time.Time{})
_ = conn.SetWriteDeadline(time.Now().Add(writeTimeout))
s.reply(conn, s.serve(req))
}
// serve runs one request and describes the outcome.
//
// Nothing bounds a command's run; a context would have to reach every Run. What is
// bounded is the silence: a command still running at slowAfter is named in the log, with
// the slot it is holding, so a socket that has gone "busy" can be traced to the commands
// that wedged it. The timer is stopped however the command ends, a panic included, so
// the line is only ever written about a command that is still running.
func (s *Server) serve(req *Request) *Response {
slow := time.AfterFunc(s.slowAfter, func() {
s.l.Warn("Local control command is still running and holding a connection slot",
"command", req.Command,
"args", req.Args,
"after", s.slowAfter,
"max", s.maxConns,
)
})
defer slow.Stop()
res := &Response{Version: ProtocolVersion}
reg := s.src
if req.Describe {
schema, err := json.Marshal(reg.Schema())
if err != nil {
res.Error = err.Error()
return res
}
res.Result = schema
return res
}
if req.Command == "" {
res.Error = "no command given"
return res
}
params, err := stringParams(req.Params)
if err != nil {
res.Error = err.Error()
return res
}
format, err := parseFormat(req.Format)
if err != nil {
res.Error = err.Error()
return res
}
cmd, err := reg.Resolve(commands.Name(req.Command))
if err != nil {
s.l.Debug("Local control command refused", "command", req.Command, "error", err)
res.Error = err.Error()
return res
}
r, err := commands.Invoke(commands.Request{Command: cmd, Params: params, Args: req.Args})
if err != nil {
s.l.Debug("Local control command failed", "command", req.Command, "error", err)
res.Error = err.Error()
return res
}
// A command that acts does so now, before the one reply this surface gives, and the
// reply carries both halves: the line Run promised and the outcome Then produced.
// In JSON only the outcome travels, since the promise is for a person to read.
var preface []byte
if cmd.Then != nil && r.OK {
if format == commands.FormatText && !r.Empty() {
buf := &bytes.Buffer{}
if err := commands.Render(cmd, r, format, buf); err == nil {
preface = buf.Bytes()
}
}
r = cmd.Then()
}
// A command with nothing to say gets an empty, successful reply.
if r.Empty() {
res.Text = preface
return res
}
return render(res, cmd, r, format, preface)
}
// render turns a command's Result into the response, in the format the caller asked for,
// through commands.Render, so the JSON is byte for byte what -json prints over ssh.
func render(res *Response, cmd *commands.Command, r commands.Result, format commands.Format, preface []byte) *Response {
buf := &bytes.Buffer{}
if err := commands.Render(cmd, r, format, buf); err != nil {
res.Error = err.Error()
return res
}
if format != commands.FormatText {
// The caller is a program. Carry the payload as JSON rather than as a string that
// happens to contain some; the newline Render ends a line with is not the value's.
res.Result = bytes.TrimSuffix(buf.Bytes(), []byte("\n"))
// A command that ran and refused is a failure to whoever asked, in either format.
// The structured outcome still travels in Result for a program to read; Error
// carries the reason so the one rule, Error set means failed, holds here too and
// a client never has to guess from the payload.
if !r.OK {
res.Error = r.Reason
}
return res
}
// A command that ran and refused reads as an error to whoever asked for text: they
// are a person or a script, and both expect a refusal on the error stream.
if !r.OK {
res.Error = string(append(preface, buf.Bytes()...))
return res
}
// The bytes the command wrote, as they are; see Response.Text.
res.Text = append(preface, buf.Bytes()...)
return res
}
// parseFormat turns the request's format into the one the registry understands. An
// unknown value is an error rather than a silent default, so a client that asks for
// something this daemon cannot do hears about it.
func parseFormat(s string) (commands.Format, error) {
switch s {
case "", "json":
return commands.FormatJSON, nil
case "text":
return commands.FormatText, nil
default:
return commands.FormatJSON, fmt.Errorf("unknown format %q, expected \"json\" or \"text\"", s)
}
}
// reply writes one response frame.
func (s *Server) reply(conn net.Conn, res *Response) {
_ = conn.SetWriteDeadline(time.Now().Add(writeTimeout))
if err := writeFrame(conn, res); err != nil {
s.l.Debug("Could not send a local control response", "error", err)
}
}
// stringParams converts JSON's types into the strings flag.Value.Set expects, so a
// caller writes true and 3 rather than "true" and "3".
func stringParams(in map[string]any) (map[string]string, error) {
if len(in) == 0 {
return nil, nil
}
out := make(map[string]string, len(in))
for k, v := range in {
switch t := v.(type) {
case bool:
out[k] = strconv.FormatBool(t)
case string:
out[k] = t
case float64:
// encoding/json gives every number as a float64; render whole ones without
// a decimal point so an int flag accepts them.
if t != math.Trunc(t) {
out[k] = strconv.FormatFloat(t, 'f', -1, 64)
continue
}
// A whole float64 can be far outside int64, and Go's conversion of one that is
// gives an unspecified value rather than an error, so the bounds are checked
// first. MinInt64 is exactly representable and is in range. MaxInt64 is not:
// as a float64 it rounds up to 2^63, which is the first value out of range,
// hence the >= on that side.
if t < math.MinInt64 || t >= math.MaxInt64 {
return nil, fmt.Errorf("parameter %q is %s, outside the range of a 64-bit integer",
k, strconv.FormatFloat(t, 'g', -1, 64))
}
out[k] = strconv.FormatInt(int64(t), 10)
case nil:
return nil, fmt.Errorf("parameter %q has no value", k)
default:
return nil, fmt.Errorf("parameter %q is not a string, number or boolean", k)
}
}
return out, nil
}
func readRequest(conn net.Conn) (*Request, error) {
req := &Request{}
if err := readFrame(conn, MaxRequestBytes, "request", req, true); err != nil {
return nil, err
}
if req.Version != ProtocolVersion {
return nil, fmt.Errorf(
"this nebula speaks local control protocol v%d, the client sent v%d; the two binaries are different versions",
ProtocolVersion, req.Version,
)
}
return req, nil
}
-950
View File
@@ -1,950 +0,0 @@
//go:build !windows
package localapi
import (
"bytes"
"context"
"encoding/json"
"errors"
"flag"
"fmt"
"io"
"math"
"net"
"os"
"path/filepath"
"strings"
"sync"
"testing"
"time"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
"golang.org/x/sys/unix"
)
// testSocket returns a short socket path whose directory does not exist yet and whose
// parent is private: one level below the test's own directory, so that listen has a
// directory of its own to create.
func testSocket(t *testing.T) string {
t.Helper()
return filepath.Join(test.ShortTempDir(t), "n", "s.sock")
}
func serve(t *testing.T, reg *commands.Registry) (*Server, string) {
t.Helper()
path := testSocket(t)
s := NewServer(test.NewLogger(), reg)
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
t.Cleanup(s.Stop)
return s, path
}
// serveAt starts a server on an exact path, for tests that care where the socket is.
func serveAt(t *testing.T, path string) (*Server, func()) {
t.Helper()
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
return s, s.Stop
}
// echoCommand reports its operands, for tests on either surface.
func echoCommand() *commands.Command {
return &commands.Command{
Name: "echo",
ShortDescription: "echoes its arguments",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Value(fmt.Sprintf("%v", a)), nil
},
}
}
func basicRegistry() *commands.Registry {
reg := commands.New()
reg.Register(echoCommand())
reg.Register(&commands.Command{
Name: "boom",
ShortDescription: "panics",
Run: func(fs any, a []string) (commands.Result, error) {
panic("this command is broken")
},
})
return reg
}
// run invokes one command and returns its output and its error text. An empty error is
// success: the protocol has no separate status field.
func run(t *testing.T, path, command string, args ...string) (string, string) {
t.Helper()
res, err := (&Client{Path: path}).Invoke(command, nil, args, "")
require.NoError(t, err)
return textOf(res), res.Error
}
// textOf is what a test reads as a command's output: the text when text was asked for,
// else the JSON result, unquoted when it is a JSON string so a test can compare words.
func textOf(res *Response) string {
if len(res.Text) > 0 {
return string(res.Text)
}
if len(res.Result) > 0 && res.Result[0] == '"' {
var s string
if err := json.Unmarshal(res.Result, &s); err == nil {
return s
}
}
return string(res.Result)
}
func TestServeAllowedCommand(t *testing.T) {
_, path := serve(t, basicRegistry())
out, errOut := run(t, path, "echo", "a", "b")
assert.Equal(t, "[a b]", out)
assert.Empty(t, errOut, "a command that worked reports no error")
}
// TestPanicInCommandDoesNotKillTheDaemon: a broken command costs a connection, not the
// overlay.
func TestPanicInCommandDoesNotKillTheDaemon(t *testing.T) {
_, path := serve(t, basicRegistry())
out, errOut := run(t, path, "boom")
assert.Empty(t, out)
assert.Contains(t, errOut, "command failed")
// The server is still serving, which is the point of the test.
out, errOut = run(t, path, "echo", "still", "here")
assert.Equal(t, "[still here]", out)
assert.Empty(t, errOut)
}
// TestSocketPermissions: the modes are pinned, not left to the umask. A 077 umask would
// make the directory 0700 and hide a root daemon's socket from an unprivileged nebula
// ctl.
func TestSocketPermissions(t *testing.T) {
for _, umask := range []int{0, 0077} {
t.Run(fmt.Sprintf("umask %04o", umask), func(t *testing.T) {
old := unix.Umask(umask)
defer unix.Umask(old)
_, path := serve(t, basicRegistry())
fi, err := os.Stat(path)
require.NoError(t, err)
assert.Equal(t, os.FileMode(0600), fi.Mode().Perm(), "socket must not be reachable by other users")
di, err := os.Stat(filepath.Dir(path))
require.NoError(t, err)
assert.Equal(t, os.FileMode(0755), di.Mode().Perm(), "a directory we create ourselves is searchable, so ServedByRoot can see into it")
})
}
}
func TestLiveSocketIsNotStolen(t *testing.T) {
first, path := serve(t, basicRegistry())
require.NotNil(t, first)
second := NewServer(test.NewLogger(), basicRegistry())
err := second.Rebind(context.Background(), path, ownerOnly())
assert.ErrorContains(t, err, "already in use", "a live endpoint must not be unlinked by a second instance")
}
// TestLiveSocketIsNotStolenBeforeItsChmod: between an instance's bind and its chmod the
// socket is at whatever mode the umask left. A second instance starting in that window
// must still see it as live, not stale: only a refused connect proves nothing is behind
// the file.
func TestLiveSocketIsNotStolenBeforeItsChmod(t *testing.T) {
first, path := serve(t, basicRegistry())
require.NotNil(t, first)
require.NoError(t, os.Chmod(path, 0666))
second := NewServer(test.NewLogger(), basicRegistry())
err := second.Rebind(context.Background(), path, ownerOnly())
require.ErrorContains(t, err, "already in use", "a live socket at any mode is not thereby dead")
_, err = os.Stat(path)
require.NoError(t, err, "the first instance's socket is still there")
require.NoError(t, os.Chmod(path, 0600))
out, _ := run(t, path, "echo", "ok")
assert.Equal(t, "[ok]", out, "and still served")
}
func TestRebindMovesTheSocket(t *testing.T) {
s, path := serve(t, basicRegistry())
next := filepath.Join(filepath.Dir(path), "n.sock")
require.NoError(t, s.Rebind(context.Background(), next, ownerOnly()))
assert.Equal(t, next, s.Path())
out, _ := run(t, next, "echo", "moved")
assert.Equal(t, "[moved]", out)
_, err := os.Stat(path)
assert.True(t, os.IsNotExist(err), "the old socket should be unlinked")
}
func TestRebindFailureKeepsTheOldSocket(t *testing.T) {
s, path := serve(t, basicRegistry())
err := s.Rebind(context.Background(), "/nonexistent-dir-for-tests/x.sock", ownerOnly())
require.Error(t, err)
assert.Equal(t, path, s.Path(), "a bad path must leave the working endpoint alone")
out, _ := run(t, path, "echo", "ok")
assert.Equal(t, "[ok]", out)
}
// TestSemaphoreSurvivesRestart pins the rule that the connection limiter is made once. A
// handler releases into the channel it acquired from, so if Rebind replaced it after a
// Stop, a command still running across the restart would block forever on the new,
// empty channel and leak its goroutine and connection.
func TestSemaphoreSurvivesRestart(t *testing.T) {
s := NewServer(test.NewLogger(), basicRegistry())
path := testSocket(t)
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
first := s.sem
require.NotNil(t, first)
s.Stop()
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
defer s.Stop()
assert.Equal(t, first, s.sem, "Rebind after Stop must keep the same semaphore")
require.NoError(t, s.Rebind(context.Background(), testSocket(t), ownerOnly()))
assert.Equal(t, first, s.sem, "Rebind must keep the same semaphore")
}
// TestRestartIsWatchedOnItsOwnContext: the ctx watcher is installed by the first bind of
// a run and cleared by Stop, so a server stopped by one context's cancellation and
// started again under another is stopped when that second context is cancelled, not
// left serving until the process exits.
func TestRestartIsWatchedOnItsOwnContext(t *testing.T) {
s := NewServer(test.NewLogger(), basicRegistry())
path := testSocket(t)
ctxA, cancelA := context.WithCancel(context.Background())
require.NoError(t, s.Rebind(ctxA, path, ownerOnly()))
cancelA()
require.Eventually(t, func() bool { return s.Path() == "" }, time.Second, 10*time.Millisecond, "the first context must stop the server")
ctxB, cancelB := context.WithCancel(context.Background())
require.NoError(t, s.Rebind(ctxB, path, ownerOnly()))
defer s.Stop()
assert.Equal(t, path, s.Path())
cancelB()
require.Eventually(t, func() bool { return s.Path() == "" }, time.Second, 10*time.Millisecond, "the second context must stop the server too")
_, err := os.Stat(path)
assert.True(t, os.IsNotExist(err), "the socket must be removed when the second context stops the server")
}
// TestStopDetachesTheWatcher: a server stopped by Stop and started again under another
// context belongs to that context alone. The one that started the earlier run, ending
// later, must not take the new run down with it.
func TestStopDetachesTheWatcher(t *testing.T) {
s := NewServer(test.NewLogger(), basicRegistry())
path := testSocket(t)
ctxA, cancelA := context.WithCancel(context.Background())
t.Cleanup(cancelA)
require.NoError(t, s.Rebind(ctxA, path, ownerOnly()))
s.Stop()
ctxB, cancelB := context.WithCancel(context.Background())
t.Cleanup(cancelB)
require.NoError(t, s.Rebind(ctxB, path, ownerOnly()))
t.Cleanup(s.Stop)
cancelA()
time.Sleep(50 * time.Millisecond)
assert.Equal(t, path, s.Path(), "the first run's context has nothing to do with the second run")
_, err := (&Client{Path: path}).Describe()
require.NoError(t, err)
cancelB()
require.Eventually(t, func() bool { return s.Path() == "" }, time.Second, 10*time.Millisecond, "the second run's context stops it")
}
func TestVersionMismatchIsExplained(t *testing.T) {
_, path := serve(t, basicRegistry())
conn, err := net.Dial("unix", path)
require.NoError(t, err)
defer conn.Close()
body, _ := json.Marshal(Request{Version: ProtocolVersion + 1, Command: "echo"})
_, err = conn.Write(append(body, '\n'))
require.NoError(t, err)
res, err := readResponse(conn)
require.NoError(t, err)
assert.Contains(t, res.Error, "different versions")
}
func TestOversizedRequestIsRejected(t *testing.T) {
_, path := serve(t, basicRegistry())
conn, err := net.Dial("unix", path)
require.NoError(t, err)
defer conn.Close()
_, err = conn.Write(append(bytes.Repeat([]byte("a"), MaxRequestBytes+10), '\n'))
if err == nil {
res, rerr := readResponse(conn)
require.NoError(t, rerr)
assert.NotEmpty(t, res.Error)
}
}
// TestConcurrentClientsAreBounded checks that going over the connection cap is answered
// rather than queued, so a burst of clients cannot pile up against the packet path.
// TestSlowCommandIsNamedInTheLog: a command that does not return holds its slot for
// good, and enough of them leave the socket answering only "busy". Both are logged, the
// command by name once it has run past slowAfter and each refused connection as it
// happens, so a busy socket can be traced to what wedged it.
func TestSlowCommandIsNamedInTheLog(t *testing.T) {
reg := commands.New()
release := make(chan struct{})
reg.Register(&commands.Command{
Name: "block",
ShortDescription: "blocks until released",
Run: func(fs any, a []string) (commands.Result, error) {
<-release
return commands.Value("done"), nil
},
})
var logs lockedBuffer
s := NewServer(test.NewLoggerWithOutput(&logs), reg)
s.maxConns = 1
s.slowAfter = 20 * time.Millisecond
path := testSocket(t)
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
defer s.Stop()
var wg sync.WaitGroup
wg.Add(1)
go func() {
defer wg.Done()
_, _ = (&Client{Path: path}).Invoke("block", nil, []string{"x"}, "")
}()
require.Eventually(t, func() bool {
return strings.Contains(logs.String(), "still running")
}, 2*time.Second, 10*time.Millisecond)
assert.Contains(t, logs.String(), "command=block")
assert.Contains(t, logs.String(), "args=[x]")
res, err := (&Client{Path: path}).Invoke("block", nil, nil, "")
require.NoError(t, err)
assert.Contains(t, res.Error, "busy")
assert.Contains(t, logs.String(), "at capacity")
close(release)
wg.Wait()
}
// TestPanickedCommandIsNotReportedAsStillRunning: the slow command line is about a slot
// that is still held. A command that panicked has given its slot back, and the client
// its answer, so no line must claim otherwise after the fact.
func TestPanickedCommandIsNotReportedAsStillRunning(t *testing.T) {
reg := commands.New()
reg.Register(&commands.Command{
Name: "boom",
ShortDescription: "panics",
Run: func(fs any, a []string) (commands.Result, error) {
panic("this command is broken")
},
})
var logs lockedBuffer
s := NewServer(test.NewLoggerWithOutput(&logs), reg)
s.slowAfter = 20 * time.Millisecond
path := testSocket(t)
require.NoError(t, s.Rebind(context.Background(), path, ownerOnly()))
defer s.Stop()
res, err := (&Client{Path: path}).Invoke("boom", nil, nil, "")
require.NoError(t, err)
assert.Equal(t, "command failed", res.Error)
time.Sleep(100 * time.Millisecond)
assert.Contains(t, logs.String(), "Recovered from panic")
assert.NotContains(t, logs.String(), "still running")
}
// lockedBuffer is a strings.Builder the server's goroutines and the test can share.
type lockedBuffer struct {
mu sync.Mutex
b strings.Builder
}
func (l *lockedBuffer) Write(p []byte) (int, error) {
l.mu.Lock()
defer l.mu.Unlock()
return l.b.Write(p)
}
func (l *lockedBuffer) String() string {
l.mu.Lock()
defer l.mu.Unlock()
return l.b.String()
}
// flakyListener fails its first few Accepts the way a process that is out of file
// descriptors does, then behaves.
type flakyListener struct {
net.Listener
failures int
}
func (f *flakyListener) Accept() (net.Conn, error) {
if f.failures > 0 {
f.failures--
return nil, &net.OpError{Op: "accept", Net: "unix", Err: errors.New("too many open files")}
}
return f.Listener.Accept()
}
// TestAcceptSurvivesTransientErrors: an Accept that fails for a reason other than the
// listener being closed is retried, so a moment at the descriptor limit does not take
// the control socket down for the rest of the process. Closing the listener is still
// what ends the loop.
func TestAcceptSurvivesTransientErrors(t *testing.T) {
path := testSocket(t)
require.NoError(t, os.Mkdir(filepath.Dir(path), 0700))
ln, err := listen(path, ownerOnly())
require.NoError(t, err)
s := NewServer(test.NewLogger(), basicRegistry())
done := make(chan struct{})
go func() {
defer close(done)
s.accept(&flakyListener{Listener: ln, failures: 3}, make(chan struct{}, 1))
}()
out, errText := run(t, path, "echo", "still", "here")
assert.Empty(t, errText)
assert.Equal(t, "[still here]", out)
require.NoError(t, ln.Close())
select {
case <-done:
case <-time.After(5 * time.Second):
t.Fatal("accept loop did not exit when the listener was closed")
}
}
func TestStopRemovesTheSocket(t *testing.T) {
s, path := serve(t, basicRegistry())
s.Stop()
_, err := os.Stat(path)
assert.True(t, os.IsNotExist(err))
}
func TestContextCancellationStops(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
path := testSocket(t)
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(ctx, path, ownerOnly()))
cancel()
assert.Eventually(t, func() bool {
_, err := os.Stat(path)
return os.IsNotExist(err)
}, 2*time.Second, 20*time.Millisecond)
}
// TestNamedParametersDriveFlags is the point of the protocol: a caller passes options by
// name, with JSON's own types, and never assembles a command line.
// TestStringParamsNumbers: JSON numbers arrive as float64. Whole ones render as integers
// so int flags accept them, and a whole one outside int64 is refused rather than
// converted, since Go's float-to-int conversion of an out-of-range value is unspecified
// and would set a number the caller never sent.
func TestStringParamsNumbers(t *testing.T) {
cases := map[string]struct {
in float64
want string
wantErr bool
}{
"whole": {in: 3, want: "3"},
"negative whole": {in: -1, want: "-1"},
"fraction": {in: 1.5, want: "1.5"},
"min int64": {in: math.MinInt64, want: "-9223372036854775808"},
"largest in range": {in: 9223372036854774784, want: "9223372036854774784"},
"2^63": {in: 9223372036854775808, wantErr: true},
"1e19": {in: 1e19, wantErr: true},
"below min int64": {in: -9.3e18, wantErr: true},
}
for name, tc := range cases {
t.Run(name, func(t *testing.T) {
out, err := stringParams(map[string]any{"n": tc.in})
if tc.wantErr {
require.Error(t, err)
require.ErrorContains(t, err, `parameter "n" is `)
assert.ErrorContains(t, err, "outside the range of a 64-bit integer")
return
}
require.NoError(t, err)
assert.Equal(t, tc.want, out["n"])
})
}
}
func TestNamedParametersDriveFlags(t *testing.T) {
reg := commands.New()
reg.Register(&commands.Command{
Name: "report",
ShortDescription: "reports its flags",
Flags: func() (*flag.FlagSet, any) {
fl := flag.NewFlagSet("", flag.ContinueOnError)
s := struct {
Verbose bool
Label string
}{}
fl.BoolVar(&s.Verbose, "verbose", false, "say more")
fl.StringVar(&s.Label, "label", "none", "a label")
return fl, &s
},
Run: func(fs any, a []string) (commands.Result, error) {
s := fs.(*struct {
Verbose bool
Label string
})
return commands.Value(fmt.Sprintf("verbose=%v label=%s args=%v", s.Verbose, s.Label, a)), nil
},
})
_, path := serve(t, reg)
res, err := (&Client{Path: path}).Invoke("report",
map[string]string{"verbose": "true", "label": "here"},
[]string{"an-operand"},
"",
)
require.NoError(t, err)
assert.Empty(t, res.Error)
assert.Equal(t, "verbose=true label=here args=[an-operand]", textOf(res))
}
// TestUnknownParameterIsReported: a misspelled parameter is an error rather than a
// silently applied default, which is the difference between an API and a guess.
func TestUnknownParameterIsReported(t *testing.T) {
reg := commands.New()
reg.Register(&commands.Command{
Name: "report",
ShortDescription: "takes a flag",
Flags: func() (*flag.FlagSet, any) {
fl := flag.NewFlagSet("", flag.ContinueOnError)
fl.Bool("verbose", false, "say more")
return fl, nil
},
Run: func(fs any, a []string) (commands.Result, error) { return commands.Nothing(), nil },
})
_, path := serve(t, reg)
res, err := (&Client{Path: path}).Invoke("report", map[string]string{"verbos": "true"}, nil, "")
require.NoError(t, err)
assert.Contains(t, res.Error, `report has no parameter "verbos"`)
}
func TestParametersOnACommandThatTakesNoneAreReported(t *testing.T) {
_, path := serve(t, basicRegistry())
res, err := (&Client{Path: path}).Invoke("echo", map[string]string{"nope": "true"}, nil, "")
require.NoError(t, err)
assert.Contains(t, res.Error, "echo takes no parameters")
}
// TestJSONOutputIsAJSONValue: a command that emits JSON lands in the response as JSON,
// not as a string a caller has to parse a second time.
func TestJSONOutputIsAJSONValue(t *testing.T) {
reg := commands.New()
reg.Register(&commands.Command{
Name: "emit",
ShortDescription: "emits json",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Value(map[string]any{"hosts": []int{1, 2, 3}}), nil
},
})
_, path := serve(t, reg)
res, err := (&Client{Path: path}).Invoke("emit", nil, nil, "")
require.NoError(t, err)
var decoded struct {
Hosts []int `json:"hosts"`
}
require.NoError(t, json.Unmarshal(res.Result, &decoded))
assert.Equal(t, []int{1, 2, 3}, decoded.Hosts)
}
// TestStringValueIsAJSONString: a command whose value is a string answers a JSON request
// with a JSON string, quotes and all.
func TestStringValueIsAJSONString(t *testing.T) {
_, path := serve(t, basicRegistry())
res, err := (&Client{Path: path}).Invoke("echo", nil, []string{"hi"}, "")
require.NoError(t, err)
var s string
require.NoError(t, json.Unmarshal(res.Result, &s))
assert.Equal(t, "[hi]", s)
}
func TestDescribeListsTheCommands(t *testing.T) {
_, path := serve(t, basicRegistry())
schema, err := (&Client{Path: path}).Describe()
require.NoError(t, err)
names := make([]string, 0, len(schema))
for _, c := range schema {
names = append(names, string(c.Name))
}
assert.ElementsMatch(t, []string{"help", "echo", "boom"}, names)
}
// TestPresentationParametersAreRefused: the schema lists json and pretty on every
// command, because the command lines take them, but a request has one way to choose its
// rendering, the format field. A client that sends them as parameters is pointed there
// rather than given a second, weaker way to say the same thing.
func TestPresentationParametersAreRefused(t *testing.T) {
reg := commands.New()
reg.Register(&commands.Command{
Name: "pair",
ShortDescription: "returns a pair",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Value(map[string]int{"a": 1}), nil
},
Text: func(v any, w io.Writer) error {
_, err := fmt.Fprintln(w, "a is one")
return err
},
})
_, path := serve(t, reg)
c := &Client{Path: path}
for _, name := range []string{"json", "pretty"} {
res, err := c.Invoke("pair", map[string]string{name: "true"}, nil, "text")
require.NoError(t, err)
assert.Contains(t, res.Error, name+" is a command line flag, not a parameter")
assert.Contains(t, res.Error, "format")
assert.Empty(t, res.Result, "a refused request runs nothing")
}
// The format field is the one that counts.
res, err := c.Invoke("pair", nil, nil, "text")
require.NoError(t, err)
require.Empty(t, res.Error)
assert.Equal(t, "a is one\n", textOf(res))
res, err = c.Invoke("pair", nil, nil, "")
require.NoError(t, err)
require.Empty(t, res.Error)
assert.JSONEq(t, `{"a":1}`, string(res.Result), "the API default is JSON")
}
// TestTextRequestToATextlessCommand: a command with no human rendering still answers a
// text request when the answer is a refusal, since a refusal always has one. Only a
// successful result has nothing to say in text, and that is reported then, by the daemon.
func TestTextRequestToATextlessCommand(t *testing.T) {
reg := commands.New()
reg.Register(&commands.Command{
Name: "tunnel",
ShortDescription: "data on success, a refusal without an operand",
Run: func(fs any, a []string) (commands.Result, error) {
if len(a) == 0 {
return commands.Failed("No vpn address was provided"), nil
}
return commands.Value(map[string]int{"index": 7}), nil
},
JSONOnly: true,
})
_, path := serve(t, reg)
c := &Client{Path: path}
res, err := c.Invoke("tunnel", nil, nil, "text")
require.NoError(t, err)
assert.Equal(t, "No vpn address was provided\n", res.Error, "the refusal is rendered, not the missing text form")
res, err = c.Invoke("tunnel", nil, []string{"10.0.0.1"}, "text")
require.NoError(t, err)
assert.Contains(t, res.Error, "tunnel has no human readable output yet; pass -json or -pretty")
assert.Empty(t, res.Result)
res, err = c.Invoke("tunnel", nil, []string{"10.0.0.1"}, "json")
require.NoError(t, err)
require.Empty(t, res.Error)
assert.JSONEq(t, `{"index":7}`, string(res.Result))
// A refusal asked for as JSON carries the outcome for a program and the reason in
// Error, so a client exits non-zero on the envelope alone, never by sniffing the
// payload for an ok field some unrelated result might also have.
res, err = c.Invoke("tunnel", nil, nil, "json")
require.NoError(t, err)
assert.Equal(t, "No vpn address was provided", res.Error)
assert.JSONEq(t, `{"ok":false,"reason":"No vpn address was provided"}`, string(res.Result))
}
// TestRebindAfterFailedBindStillStopsOnCancel covers a daemon whose initial listen
// failed and which later Rebinds onto a good path from a reloaded config. The context
// watcher has to be installed by that Rebind, or the socket outlives the daemon.
func TestRebindAfterFailedBindStillStopsOnCancel(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
defer cancel()
base := test.ShortTempDir(t)
// listen creates one level only, so a socket directory under a parent that is not
// there yet fails to bind, the shape of a runtime directory that does not exist.
s := NewServer(test.NewLogger(), basicRegistry())
require.Error(t, s.Rebind(ctx, filepath.Join(base, "missing", "n", "s.sock"), ownerOnly()))
good := filepath.Join(base, "n", "s.sock")
require.NoError(t, s.Rebind(ctx, good, ownerOnly()))
t.Cleanup(s.Stop)
_, err := (&Client{Path: good}).Describe()
require.NoError(t, err)
cancel()
assert.Eventually(t, func() bool {
_, err := os.Stat(good)
return os.IsNotExist(err)
}, 2*time.Second, 20*time.Millisecond)
}
// TestCommandRegisteredAfterStartIsServed: the server filters the registry per request,
// so a command an embedder registers after wiring the socket is served, and described,
// just as it would be over ssh.
func TestCommandRegisteredAfterStartIsServed(t *testing.T) {
reg := basicRegistry()
_, path := serve(t, reg)
reg.Register(&commands.Command{
Name: "late",
ShortDescription: "registered after the server started",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Value("late"), nil
},
})
c := &Client{Path: path}
schema, err := c.Describe()
require.NoError(t, err)
names := make([]string, 0, len(schema))
for _, cs := range schema {
names = append(names, string(cs.Name))
}
assert.Contains(t, names, "late")
res, err := c.Invoke("late", nil, nil, "")
require.NoError(t, err)
require.Empty(t, res.Error)
assert.JSONEq(t, `"late"`, string(res.Result))
}
// TestThenIsAnsweredInTheOneReply: this surface answers in one piece, so a command that
// acts does so before the reply, and the reply carries Run's line and Then's outcome in
// text, and the outcome alone in JSON.
func TestThenIsAnsweredInTheOneReply(t *testing.T) {
reg := commands.New()
acted := false
reg.Register(&commands.Command{
Name: "act",
ShortDescription: "replies, then acts",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Did("about to act"), nil
},
Then: func() commands.Result {
acted = true
return commands.Did("acted")
},
})
_, path := serve(t, reg)
res, err := (&Client{Path: path}).Invoke("act", nil, nil, "text")
require.NoError(t, err)
assert.True(t, acted)
assert.Empty(t, res.Error)
assert.Equal(t, "about to act\nacted\n", string(res.Text))
res, err = (&Client{Path: path}).Invoke("act", nil, nil, "")
require.NoError(t, err)
assert.JSONEq(t, `{"ok":true,"reason":"acted"}`, string(res.Result))
// A refusal from Then still carries Run's line ahead of it in text, as ssh prints it.
acted = false
reg.Register(&commands.Command{
Name: "fail",
ShortDescription: "replies, then fails to act",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Did("about to act"), nil
},
Then: func() commands.Result {
return commands.Failed("could not act")
},
})
res, err = (&Client{Path: path}).Invoke("fail", nil, nil, "text")
require.NoError(t, err)
assert.Equal(t, "about to act\ncould not act\n", res.Error)
assert.Empty(t, res.Text)
// A Run that refused is the outcome, and Then does not run.
acted = false
reg.Register(&commands.Command{
Name: "refuse",
ShortDescription: "refuses, so never acts",
Run: func(fs any, a []string) (commands.Result, error) {
return commands.Failed("not like that"), nil
},
Then: func() commands.Result {
acted = true
return commands.Did("acted anyway")
},
})
res, err = (&Client{Path: path}).Invoke("refuse", nil, nil, "text")
require.NoError(t, err)
assert.Equal(t, "not like that\n", res.Error)
assert.False(t, acted)
}
// TestRebindRefusesDuringShutdown: a reload that lands after the daemon's context is
// cancelled must not bring the endpoint back. Stop is not terminal, so Rebind has to
// take the answer from the context rather than from the running flag.
func TestRebindRefusesDuringShutdown(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
defer cancel()
// Built by hand rather than through serve: the context watcher is installed once,
// by the first bind, and this test needs it on a context it can cancel.
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(ctx, testSocket(t), ownerOnly()))
t.Cleanup(s.Stop)
cancel()
assert.Eventually(t, func() bool { return s.Path() == "" }, 2*time.Second, 20*time.Millisecond, "the watcher stops the server")
next := testSocket(t)
err := s.Rebind(ctx, next, ownerOnly())
require.Error(t, err)
require.ErrorContains(t, err, "shutting down")
assert.Empty(t, s.Path(), "the server stays stopped")
_, statErr := os.Stat(next)
assert.True(t, os.IsNotExist(statErr), "and the refused socket is not left on disk")
}
// TestRebindAndStopSerialize hammers Rebind against Stop. Whichever runs last decides
// the final state, so after the dust settles a Stop must leave nothing listening and no
// socket file behind, on either path.
func TestRebindAndStopSerialize(t *testing.T) {
a := testSocket(t)
b := testSocket(t)
for i := 0; i < 50; i++ {
s := NewServer(test.NewLogger(), basicRegistry())
require.NoError(t, s.Rebind(context.Background(), a, ownerOnly()))
var wg sync.WaitGroup
wg.Add(2)
go func() {
defer wg.Done()
_ = s.Rebind(context.Background(), b, ownerOnly())
}()
go func() {
defer wg.Done()
s.Stop()
}()
wg.Wait()
// The Rebind may have run after the Stop, which restarts the server by design.
// A final Stop then has to take everything down.
s.Stop()
assert.Empty(t, s.Path())
for _, p := range []string{a, b} {
_, err := os.Stat(p)
assert.True(t, os.IsNotExist(err), "iteration %d: %s left behind", i, p)
}
}
}
// TestUnknownRequestFieldIsRefused: a request with a field the protocol does not define
// is answered with the field named, the way an unknown parameter is, and not run.
func TestUnknownRequestFieldIsRefused(t *testing.T) {
_, path := serve(t, basicRegistry())
conn, err := dial(path, time.Second)
require.NoError(t, err)
defer conn.Close()
require.NoError(t, writeFrame(conn, map[string]any{"version": ProtocolVersion, "command": "echo", "arguments": []string{"x"}}))
var res Response
require.NoError(t, readFrame(conn, MaxResponseBytes, "response", &res, false))
assert.Contains(t, res.Error, `unknown field "arguments"`)
assert.Empty(t, res.Result)
}
// TestTrailingBytesOnARequestAreRefused: a request line is one object and nothing
// more; a second object or stray bytes after it are refused rather than ignored.
func TestTrailingBytesOnARequestAreRefused(t *testing.T) {
_, path := serve(t, basicRegistry())
for _, line := range []string{
`{"version":1,"command":"echo"} junk` + "\n",
`{"version":1,"command":"echo"}{"version":1,"command":"boom"}` + "\n",
} {
conn, err := dial(path, time.Second)
require.NoError(t, err)
_, err = conn.Write([]byte(line))
require.NoError(t, err)
var res Response
require.NoError(t, readFrame(conn, MaxResponseBytes, "response", &res, false))
assert.Contains(t, res.Error, "unexpected data after the object", line)
conn.Close()
}
}
// TestTextOutputKeepsTheBytes: a text rendering travels as the bytes the command wrote.
// A JSON string would have had to be UTF-8, and a byte that is not would have come out
// as U+FFFD, where the ssh server passes it through.
func TestTextOutputKeepsTheBytes(t *testing.T) {
reg := commands.New()
raw := []byte{'n', 'a', 'm', 'e', ' ', 0xff, 0xfe, '\n'}
reg.Register(&commands.Command{
Name: "raw",
ShortDescription: "writes what a certificate held",
Run: func(fs any, a []string) (commands.Result, error) { return commands.Value("raw"), nil },
Text: func(v any, w io.Writer) error {
_, err := w.Write(raw)
return err
},
})
_, path := serve(t, reg)
res, err := (&Client{Path: path}).Invoke("raw", nil, nil, "text")
require.NoError(t, err)
assert.Empty(t, res.Error)
assert.Equal(t, raw, res.Text)
assert.Empty(t, res.Result, "text is not also carried as JSON")
}
-144
View File
@@ -1,144 +0,0 @@
package localapi
import (
"context"
"errors"
"log/slog"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/config"
)
// Wire brings up the local control socket for a running nebula, unless the config turns
// it off, and keeps it following the config across reloads.
//
// Every entry point that runs nebula as a daemon calls this, so the endpoint behaves the
// same whether nebula was started from a shell or by a service manager. It deliberately
// does not live in nebula.Main: the endpoint is on by default, so an embedder inside an
// app sandbox would have to know to turn off a socket it cannot use. Wiring it at the
// daemon entry points means embedders get the command registry and no listener.
//
// Every failure below is a warning, bar one. Nebula is expected to keep forwarding
// packets whether or not anyone can reach it; but a name that another process holds is
// not a misconfiguration, it is a process that will be answering nebula ctl in the
// daemon's place, and that is logged as an error.
//
// Call it after Control.Start, so the listener only exists once the interface is up.
func Wire(ctx context.Context, l *slog.Logger, c *config.C, reg *commands.Registry) *Server {
l = l.With("subsystem", "localapi")
srv := NewServer(l, reg)
// Registered before the first attempt, so that a path or bind that is bad at startup
// can be fixed with a config edit and a reload, the same as one that goes bad later,
// and so that a HUP landing in between is harmless: Rebind on a server already at
// the path is a no-op.
reconcile := followControlConfig(ctx, l, c, srv)
reconcile(c)
if !ControlEnabled(c) {
l.Info("Local control socket is disabled", "key", CtlEnabledKey)
}
return srv
}
// followControlConfig registers the reload callback that keeps the socket where
// local_api.control.socket says it is, moving it when the key changes and bringing one
// up while there is none, and returns it for the first attempt to call.
//
// The key is reloadable because the client finds the socket by reading the config: if the
// two could disagree, the file on disk would name a socket that is not there.
//
// Every reload is reconciled rather than gated on the key having changed. Rebind on a
// server already at the path is a no-op, so that costs nothing; and a bind that failed
// for a reason outside the config, a runtime directory that was not there yet, is fixed
// outside the config too, so the reload that follows the fix is a plain one. That holds
// for a move as much as for the first bind: a daemon left on its old path by a move that
// could not bind is one the client, reading the new path from the same file, cannot
// reach, and the operator's fix is the directory and a reload, not an edit to an
// unrelated line to make the key read as changed.
//
// The callback recovers from its own panics rather than relying on the config package,
// whose callback loop has no isolation of its own. A panic there would take the daemon
// down and skip every callback queued behind this one.
func followControlConfig(ctx context.Context, l *slog.Logger, c *config.C, srv *Server) func(*config.C) {
reconcile := func(c *config.C) {
if !ControlEnabled(c) {
if srv.Path() != "" {
srv.Stop()
l.Info("Stopped the local control socket, it is disabled in the reloaded config")
}
return
}
want, err := SocketPathFromConfig(c)
if err != nil {
warnUnusable(l, "local control socket", srv.Path(), err)
return
}
acc, err := controlAccessFrom(c)
if err != nil {
warnUnusable(l, "local control socket", srv.Path(), err)
return
}
old := srv.Path()
err = srv.Rebind(ctx, want, acc)
reportRebind(l, old, want, err, srv.Path())
}
c.RegisterReloadCallback(func(c *config.C, _ *config.Reload) {
defer func() {
if r := recover(); r != nil {
l.Error("Recovered from panic reloading the local control socket", "panic", r)
}
}()
reconcile(c)
})
return reconcile
}
// warnUnusable logs a config the endpoint cannot follow: with nothing serving, the
// endpoint is not available; with something serving, it is kept as it is.
func warnUnusable(l *slog.Logger, what, current string, err error) {
if current == "" {
l.Warn("The "+what+" is not available, its config is unusable", "error", err)
return
}
l.Warn("Ignoring unusable "+what+" config from the reload, keeping the current endpoint", "error", err, "current", current)
}
// reportRebind logs what a Rebind did beyond the listening line the bind itself writes:
// nothing for a server left where it was, a move with both paths, and a failure with
// what the operator is left with. old is the path served before the call, "" for none,
// want the one asked for, and now the path served after it: on unix a failed move keeps
// the old socket, and now is old; on windows a change of access closes the pipe before
// creating it again, and a failure there leaves nothing serving, so now is "".
//
// Nearly every failure is the operator's environment, a missing runtime directory or a
// bad path, and a warning; a name held by another process is an error, since nothing
// else will say that something is sitting where the daemon should be.
func reportRebind(l *slog.Logger, old, want string, err error, now string) {
level := l.Warn
if errors.Is(err, errEndpointHeld) {
level = l.Error
}
switch {
case err == nil:
if old != "" && old != want {
l.Info("Moved the local control socket", "from", old, "to", want)
}
case old == "":
level("Could not start the local control socket, control commands will not be available until a reload binds it", "error", err, "path", want)
default:
level("Could not bind the local control socket at the requested path", "error", err, "requested", want, "serving", now)
}
}
// errEndpointHeld is returned by listen when the endpoint's name is already held by
// some other process, which nebula will not serve alongside: whoever holds it would be
// answering nebula ctl. It is only raised on windows, where the default name can be
// held by another administrator's process and an override outside the protected
// namespace by any account's; on unix a path that is still being served is an earlier
// nebula, which clearStale reports as in use.
var errEndpointHeld = errors.New("the local control socket's name is held by another process")
-175
View File
@@ -1,175 +0,0 @@
//go:build !windows
package localapi
import (
"context"
"fmt"
"os"
"path/filepath"
"strings"
"testing"
"github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/test"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestWireRecoversFromBadPathAtStartup covers a socket path that is unusable when nebula
// starts. The operator's fix is a config edit and a reload, so the reload has to be able
// to bring up a socket that never existed, not only move one that does.
func TestWireRecoversFromBadPathAtStartup(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
t.Cleanup(cancel)
c := config.NewC(test.NewLogger())
require.NoError(t, c.LoadString(fmt.Sprintf("local_api:\n control:\n socket: /%s\n", strings.Repeat("x", sunPathMax))))
srv := Wire(ctx, test.NewLogger(), c, basicRegistry())
t.Cleanup(srv.Stop)
assert.Empty(t, srv.Path(), "an over-long path must not produce a socket")
good := testSocket(t)
require.NoError(t, c.ReloadConfigString(fmt.Sprintf("local_api:\n control:\n socket: %s\n", good)).Err)
assert.Equal(t, good, srv.Path())
out, errText := run(t, good, "echo", "hi")
assert.Empty(t, errText)
assert.Equal(t, "[hi]", out)
}
// TestRebindAfterReloadBoundTheSocket covers a HUP that lands inside Wire, after the
// reload callback is registered and before the first attempt runs. The callback does the
// first bind; the attempt must then find the socket it was about to create already up
// and leave it, not fail or rebind it, which would have Wire warn that commands are
// unavailable while they are.
func TestRebindAfterReloadBoundTheSocket(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
t.Cleanup(cancel)
path := testSocket(t)
cfg := fmt.Sprintf("local_api:\n control:\n socket: %s\n", path)
c := config.NewC(test.NewLogger())
require.NoError(t, c.LoadString(cfg))
srv := NewServer(test.NewLogger(), basicRegistry())
t.Cleanup(srv.Stop)
// Wire's first half, then the HUP.
followControlConfig(ctx, test.NewLogger(), c, srv)
require.NoError(t, c.ReloadConfigString(cfg).Err)
require.Equal(t, path, srv.Path(), "the reload callback does the first bind")
// Wire's second half.
require.NoError(t, srv.Rebind(ctx, path, ownerOnly()))
assert.Equal(t, path, srv.Path())
out, errText := run(t, path, "echo", "hi")
assert.Empty(t, errText)
assert.Equal(t, "[hi]", out)
}
// TestWireRetriesAFailedMoveOnPlainReload covers a move whose bind fails for a reason
// outside the config: the key names a path under a runtime directory that is not there
// yet. The daemon keeps the old socket, but the client reads the new path from the same
// file and cannot reach it, so the operator creates the directory and reloads. The key
// has not changed since the move was asked for, and the reload still has to make it.
func TestWireRetriesAFailedMoveOnPlainReload(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
t.Cleanup(cancel)
section := func(path string) string {
return fmt.Sprintf("local_api:\n control:\n socket: %s\n", path)
}
first := testSocket(t)
c := config.NewC(test.NewLogger())
require.NoError(t, c.LoadString(section(first)))
srv := Wire(ctx, test.NewLogger(), c, basicRegistry())
t.Cleanup(srv.Stop)
require.Equal(t, first, srv.Path())
runtimeDir := filepath.Join(privateBase(t), "run")
moved := filepath.Join(runtimeDir, DirName, "s.sock")
require.NoError(t, c.ReloadConfigString(section(moved)).Err)
assert.Equal(t, first, srv.Path(), "a move that cannot bind keeps the socket it had")
require.NoError(t, os.Mkdir(runtimeDir, 0700))
require.NoError(t, c.ReloadConfigString(section(moved)).Err)
require.False(t, c.HasChanged(CtlSocketKey), "the config is the same; the fix was on disk")
assert.Equal(t, moved, srv.Path())
_, err := os.Stat(first)
assert.True(t, os.IsNotExist(err), "the old socket is given up once the move binds")
out, errText := run(t, moved, "echo", "hi")
assert.Empty(t, errText)
assert.Equal(t, "[hi]", out)
}
// TestWireRecoversOnPlainReloadAfterBadBind covers a path that is fine but a bind that
// is not: the runtime directory is missing when nebula starts. The fix is made outside
// the config, so the config does not change; a plain reload still has to bring the
// socket up.
func TestWireRecoversOnPlainReloadAfterBadBind(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
t.Cleanup(cancel)
base := privateBase(t)
runtimeDir := filepath.Join(base, "run")
path := filepath.Join(runtimeDir, DirName, "s.sock")
cfg := fmt.Sprintf("local_api:\n control:\n socket: %s\n", path)
c := config.NewC(test.NewLogger())
require.NoError(t, c.LoadString(cfg))
srv := Wire(ctx, test.NewLogger(), c, basicRegistry())
t.Cleanup(srv.Stop)
assert.Empty(t, srv.Path(), "listen creates one level only, so a missing runtime dir fails the bind")
require.NoError(t, os.Mkdir(runtimeDir, 0700))
require.NoError(t, c.ReloadConfigString(cfg).Err)
require.False(t, c.HasChanged(CtlSocketKey), "the config is the same; the fix was on disk")
assert.Equal(t, path, srv.Path())
out, errText := run(t, path, "echo", "hi")
assert.Empty(t, errText)
assert.Equal(t, "[hi]", out)
// And a reload with a socket up and no change in the key is still a no-op.
require.NoError(t, c.ReloadConfigString(cfg).Err)
assert.Equal(t, path, srv.Path())
out, errText = run(t, path, "echo", "still")
assert.Empty(t, errText)
assert.Equal(t, "[still]", out)
}
// TestWireControlCanBeDisabled: local_api.control.enabled false serves no socket, and
// the key follows reloads both ways.
func TestWireControlCanBeDisabled(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
t.Cleanup(cancel)
path := testSocket(t)
section := func(enabled bool) string {
return fmt.Sprintf("local_api:\n control:\n enabled: %t\n socket: %s\n", enabled, path)
}
c := config.NewC(test.NewLogger())
require.NoError(t, c.LoadString(section(false)))
srv := Wire(ctx, test.NewLogger(), c, basicRegistry())
t.Cleanup(srv.Stop)
assert.Empty(t, srv.Path())
_, err := os.Lstat(path)
assert.True(t, os.IsNotExist(err), "no socket is bound")
require.NoError(t, c.ReloadConfigString(section(true)).Err)
assert.Equal(t, path, srv.Path())
out, errText := run(t, path, "echo", "on")
assert.Empty(t, errText)
assert.Equal(t, "[on]", out)
require.NoError(t, c.ReloadConfigString(section(false)).Err)
assert.Empty(t, srv.Path())
_, err = os.Lstat(path)
assert.True(t, os.IsNotExist(err), "the socket is removed")
}
-31
View File
@@ -1,31 +0,0 @@
package logging
import "log/slog"
// RateLimiter is what RateLimited needs from a rate limiter, such as a *ratelimit.Limiter.
// Allow reports whether this call may proceed, and how many calls were denied since the last one that did.
type RateLimiter interface {
Allow() (denied uint64, ok bool)
}
// RateLimited checks lim before anything is built.
// When allowed it returns l, with a "suppressed" count attached if lim denied calls since its last allowed one.
// A nil lim allows everything.
//
// if l, ok := logging.RateLimited(f.l, f.udpTxErrLimiter); ok {
// hostinfo.logger(l).Error("Failed to write outgoing packet", "error", err)
// }
func RateLimited(l *slog.Logger, lim RateLimiter) (*slog.Logger, bool) {
if lim == nil {
return l, true
}
denied, ok := lim.Allow()
if !ok {
return nil, false
}
if denied > 0 {
return l.With("suppressed", denied), true
}
return l, true
}
-106
View File
@@ -1,106 +0,0 @@
package logging
import (
"bytes"
"errors"
"io"
"log/slog"
"net/netip"
"strings"
"testing"
"time"
"github.com/slackhq/nebula/ratelimit"
)
type fakeLimiter struct {
denied uint64
ok bool
}
func (f fakeLimiter) Allow() (uint64, bool) { return f.denied, f.ok }
func TestRateLimited(t *testing.T) {
var buf bytes.Buffer
base := NewLogger(&buf)
if l, ok := RateLimited(base, fakeLimiter{ok: false}); ok || l != nil {
t.Fatalf("denied call: ok=%v logger=%v", ok, l)
}
if l, ok := RateLimited(base, fakeLimiter{ok: true}); !ok || l != base {
t.Fatal("nothing denied should hand back the same logger")
}
if l, ok := RateLimited(base, nil); !ok || l != base {
t.Fatal("a nil limiter should allow with the same logger")
}
l, ok := RateLimited(base, fakeLimiter{denied: 7, ok: true})
if !ok {
t.Fatal("allowed call denied")
}
l.Info("hello", "k", "v")
if !strings.Contains(buf.String(), "suppressed=7") {
t.Fatalf("want suppressed=7, got %q", buf.String())
}
}
// The benchmarks compare a log line through slog directly against the same line behind RateLimited: logged with
// nothing denied, denied, and logged after denials.
var (
benchErr = errors.New("sendto: no buffer space available")
benchAddr = netip.MustParseAddrPort("192.0.2.1:4242")
)
func benchLogger() *slog.Logger {
return NewLogger(io.Discard).With("localIndex", 1234, "remoteIndex", 5678)
}
func BenchmarkRateLimitedBaseline(b *testing.B) {
lg := benchLogger()
b.ReportAllocs()
for i := 0; i < b.N; i++ {
lg.Error("Failed to write outgoing packet", "error", benchErr, "udpAddr", benchAddr)
}
}
func BenchmarkRateLimitedAllowed(b *testing.B) {
lg, lim := benchLogger(), mustLimiter(b, 1, 0)
b.ReportAllocs()
for i := 0; i < b.N; i++ {
if l, ok := RateLimited(lg, lim); ok {
l.Error("Failed to write outgoing packet", "error", benchErr, "udpAddr", benchAddr)
}
}
}
func BenchmarkRateLimitedDenied(b *testing.B) {
lg, lim := benchLogger(), mustLimiter(b, 1, time.Hour)
lim.Allow()
b.ReportAllocs()
for i := 0; i < b.N; i++ {
if l, ok := RateLimited(lg, lim); ok {
l.Error("Failed to write outgoing packet", "error", benchErr, "udpAddr", benchAddr)
}
}
}
func BenchmarkRateLimitedAfterDenied(b *testing.B) {
lg, lim := benchLogger(), fakeLimiter{denied: 5, ok: true}
b.ReportAllocs()
for i := 0; i < b.N; i++ {
if l, ok := RateLimited(lg, lim); ok {
l.Error("Failed to write outgoing packet", "error", benchErr, "udpAddr", benchAddr)
}
}
}
func mustLimiter(b *testing.B, burst int, every time.Duration) *ratelimit.Limiter {
l, err := ratelimit.New(burst, every)
if err != nil {
b.Fatal(err)
}
return l
}
+7 -25
View File
@@ -12,7 +12,6 @@ import (
"strings" "strings"
"time" "time"
"github.com/slackhq/nebula/commands"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/cpupick" "github.com/slackhq/nebula/cpupick"
"github.com/slackhq/nebula/noiseutil" "github.com/slackhq/nebula/noiseutil"
@@ -72,18 +71,14 @@ func Main(c *config.C, configTest bool, buildVersion string, l *slog.Logger, dev
} }
l.Info("Firewall started", "firewallHashes", fw.GetRuleHashes()) l.Info("Firewall started", "firewallHashes", fw.GetRuleHashes())
// The command registry is shared by every surface that exposes these commands, so a ssh, err := sshd.NewSSHServer(ctx, l.With("subsystem", "sshd"))
// command cannot reach one of them without reaching the others.
cmdRegistry := commands.New()
ssh, err := sshd.NewSSHServer(ctx, l.With("subsystem", "sshd"), cmdRegistry)
if err != nil { if err != nil {
return nil, util.ContextualizeIfNeeded("Error while creating SSH server", err) return nil, util.ContextualizeIfNeeded("Error while creating SSH server", err)
} }
wireSSHReload(l, ssh, c) wireSSHReload(l, ssh, c)
var sshStart func() var sshStart func()
if c.GetBool("sshd.enabled", false) { if c.GetBool("sshd.enabled", false) {
sshStart, err = configSSH(l, ssh, c, nil) sshStart, err = configSSH(l, ssh, c)
if err != nil { if err != nil {
l.Warn("Failed to configure sshd, ssh debugging will not be available", "error", err) l.Warn("Failed to configure sshd, ssh debugging will not be available", "error", err)
sshStart = nil sshStart = nil
@@ -202,7 +197,7 @@ func Main(c *config.C, configTest bool, buildVersion string, l *slog.Logger, dev
if err != nil { if err != nil {
return nil, util.NewContextualError("Failed to open udp listener", m{"queue": i}, err) return nil, util.NewContextualError("Failed to open udp listener", m{"queue": i}, err)
} }
udpServer.ReloadConfig(c, nil) udpServer.ReloadConfig(c)
udpConns[i] = udpServer udpConns[i] = udpServer
// If port is dynamic, discover it before the next pass through the for loop // If port is dynamic, discover it before the next pass through the for loop
@@ -305,7 +300,6 @@ func Main(c *config.C, configTest bool, buildVersion string, l *slog.Logger, dev
} }
var ifce *Interface var ifce *Interface
var emitStats func()
if !configTest { if !configTest {
ifce, err = NewInterface(ctx, ifConfig) ifce, err = NewInterface(ctx, ifConfig)
if err != nil { if err != nil {
@@ -316,16 +310,6 @@ func Main(c *config.C, configTest bool, buildVersion string, l *slog.Logger, dev
lightHouse.ifce = ifce lightHouse.ifce = ifce
ifce.RegisterConfigChangeCallbacks(c) ifce.RegisterConfigChangeCallbacks(c)
// These are only read at startup
c.RegisterReloadCallback(func(c *config.C, r *config.Reload) {
r.RestartRequiredIfChanged(
"routines", "tun.routines", "listen.routines", "firewall.conntrack.routine_cache_timeout",
"listen.port", "listen.host", "stats.message_metrics",
"handshakes.try_interval", "handshakes.retries", "handshakes.trigger_buffer",
"tun.drop_local_broadcast", "tun.drop_multicast",
)
r.RestartRequiredIfChanged(osStartupOnlyKeys...)
})
ifce.reloadDisconnectInvalid(c) ifce.reloadDisconnectInvalid(c)
ifce.reloadSendRecvError(c) ifce.reloadSendRecvError(c)
ifce.reloadAcceptRecvError(c) ifce.reloadAcceptRecvError(c)
@@ -334,12 +318,9 @@ func Main(c *config.C, configTest bool, buildVersion string, l *slog.Logger, dev
go handshakeManager.Run(ctx) go handshakeManager.Run(ctx)
punchy.Start(ctx, ifce, hostMap, lightHouse) punchy.Start(ctx, ifce, hostMap, lightHouse)
// On the stats server's loop, so the gauges follow the interval it accepted
emitStats = ifce.statsEmitter()
} }
stats, err := newStatsServerFromConfig(ctx, l, c, buildVersion, configTest, emitStats) stats, err := newStatsServerFromConfig(ctx, l, c, buildVersion, configTest)
if err != nil { if err != nil {
return nil, util.ContextualizeIfNeeded("Failed to start stats emitter", err) return nil, util.ContextualizeIfNeeded("Failed to start stats emitter", err)
} }
@@ -348,14 +329,15 @@ func Main(c *config.C, configTest bool, buildVersion string, l *slog.Logger, dev
return nil, nil return nil, nil
} }
attachCommands(l, c, cmdRegistry, ifce) go ifce.emitStats(ctx, c.GetDuration("stats.interval", time.Second*10))
attachCommands(l, c, ssh, ifce)
networkChanges := udp.NewNetworkChangeMonitor(ctx, l, c) networkChanges := udp.NewNetworkChangeMonitor(ctx, l, c)
return &Control{ return &Control{
state: StateReady, state: StateReady,
f: ifce, f: ifce,
commands: cmdRegistry,
l: l, l: l,
ctx: ctx, ctx: ctx,
cancel: cancel, cancel: cancel,
-6
View File
@@ -1,6 +0,0 @@
//go:build linux && !android && !e2e_testing
package nebula
// osStartupOnlyKeys are startup only keys that only matter on linux: udp batch size, udp offloads and thread pinning
var osStartupOnlyKeys = []string{"listen.batch", "listen.udp_offloads", "tun.pin_threads", "tun.pin_threads_key", "tun.cpu_affinity"}
-6
View File
@@ -1,6 +0,0 @@
//go:build !linux || android || e2e_testing
package nebula
// osStartupOnlyKeys is empty, none of the OS specific startup keys matter here
var osStartupOnlyKeys []string
+33 -38
View File
@@ -160,7 +160,9 @@ func (f *Interface) readOutsidePackets(via ViaSender, packet []byte, rxc *rxCont
case header.TestReply: case header.TestReply:
// No-op, useful for the Roaming and connectionManager side-effects above // No-op, useful for the Roaming and connectionManager side-effects above
case header.TestRequest: case header.TestRequest:
if header.MaxOverhead+len(out) > len(rxc.scratch) { const maxCipherOverhead = 16 //todo we use this too often, needs a real importable const
const maxOverhead = header.Len + header.Len + maxCipherOverhead + maxCipherOverhead
if maxOverhead+len(out) > len(rxc.scratch) {
// A reply that cannot fit in scratch is dropped no matter the log level. // A reply that cannot fit in scratch is dropped no matter the log level.
if f.l.Enabled(context.Background(), slog.LevelDebug) { if f.l.Enabled(context.Background(), slog.LevelDebug) {
hostinfo.logger(f.l).Debug("dropping oversized test request", "payloadLen", len(out), "from", via) hostinfo.logger(f.l).Debug("dropping oversized test request", "payloadLen", len(out), "from", via)
@@ -261,16 +263,11 @@ func (f *Interface) handleOutsideRelayPacket(hostinfo *HostInfo, via ViaSender,
// closeTunnel closes a tunnel locally, it does not send a closeTunnel packet to the remote // closeTunnel closes a tunnel locally, it does not send a closeTunnel packet to the remote
func (f *Interface) closeTunnel(hostInfo *HostInfo) { func (f *Interface) closeTunnel(hostInfo *HostInfo) {
// Hold the hostmap lock across both the delete and the lighthouse eviction, final := f.hostMap.DeleteHostInfo(hostInfo)
// to avoid a race with handshake completion. if final {
f.hostMap.Lock() // We no longer have any tunnels with this vpn addr, clear learned lighthouse state to lower memory usage
defer f.hostMap.Unlock() f.lightHouse.DeleteVpnAddrs(hostInfo.vpnAddrs)
if !f.hostMap.unlockedDeleteHostInfo(hostInfo) {
return
} }
// We no longer have any tunnels with this vpn addr, clear learned lighthouse state to lower memory usage
f.lightHouse.DeleteVpnAddrs(hostInfo.vpnAddrs)
} }
// sendCloseTunnel is a helper function to send a proper close tunnel packet to a remote // sendCloseTunnel is a helper function to send a proper close tunnel packet to a remote
@@ -364,31 +361,29 @@ func parseV6(data []byte, incoming bool, fp *firewall.ParsedPacket) error {
fp.Fragment = isFragment fp.Fragment = isFragment
fp.FragAny = anyFragment fp.FragAny = anyFragment
fp.IPHdrLen = offset fp.IPHdrLen = offset
// fp is reused, so the ports are cleared for the packets that have none: fragments, protocols Nebula
// doesn't inspect, and ICMPv6 other than echo. Only the cases below set them.
fp.RemotePort, fp.LocalPort = 0, 0
if isFragment { if isFragment {
// Non-first fragments carry no transport header, so we have no ports to read // Non-first fragments carry no transport header, so we have no ports to read
fp.RemotePort = 0
fp.LocalPort = 0
return nil return nil
} }
switch proto { switch {
case iputil.IPProtocolICMPv6: case proto == iputil.IPProtocolICMPv6:
// An ICMPv6 message is at least type, code and checksum, 4 bytes. Only echo carries more than we read. // An ICMPv6 message is at least type, code and checksum, 4 bytes. Only echo carries more than we read.
if dataLen < offset+4 { if dataLen < offset+4 {
return ErrIPv6PacketTooShort return ErrIPv6PacketTooShort
} }
fp.LocalPort = 0 //incoming vs outgoing doesn't matter for icmpv6
switch data[offset] { //icmp type switch data[offset] { //icmp type
case iputil.ICMPv6TypeEchoRequest, iputil.ICMPv6TypeEchoReply: case iputil.ICMPv6TypeEchoRequest, iputil.ICMPv6TypeEchoReply:
if dataLen < offset+6 { if dataLen < offset+6 {
return ErrIPv6PacketTooShort return ErrIPv6PacketTooShort
} }
fp.RemotePort = binary.BigEndian.Uint16(data[offset+4 : offset+6]) //identifier fp.RemotePort = binary.BigEndian.Uint16(data[offset+4 : offset+6]) //identifier, orientation doesn't matter
default:
fp.RemotePort = 0
} }
case iputil.IPProtocolTCP, iputil.IPProtocolUDP: case iputil.HasPorts(proto):
if dataLen < offset+4 { if dataLen < offset+4 {
return ErrIPv6PacketTooShort return ErrIPv6PacketTooShort
} }
@@ -399,11 +394,6 @@ func parseV6(data []byte, incoming bool, fp *firewall.ParsedPacket) error {
fp.LocalPort = binary.BigEndian.Uint16(data[offset : offset+2]) fp.LocalPort = binary.BigEndian.Uint16(data[offset : offset+2])
fp.RemotePort = binary.BigEndian.Uint16(data[offset+2 : offset+4]) fp.RemotePort = binary.BigEndian.Uint16(data[offset+2 : offset+4])
} }
default:
// don't set ports for protocols Nebula doesn't inspect
fp.RemotePort = 0
fp.LocalPort = 0
} }
return nil return nil
@@ -456,18 +446,25 @@ func parseV4(data []byte, incoming bool, fp *firewall.ParsedPacket) error {
fp.RemoteAddr, _ = netip.AddrFromSlice(data[16:20]) fp.RemoteAddr, _ = netip.AddrFromSlice(data[16:20])
} }
// fp is reused, so the ports are cleared for the packets that have none: fragments, and protocols Nebula
// doesn't inspect, whose first 4 bytes (GRE, IPIP, AH, etc.) are not ports. Only the cases below set them.
fp.RemotePort, fp.LocalPort = 0, 0
if fp.Fragment { if fp.Fragment {
fp.RemotePort = 0 return nil
fp.LocalPort = 0 }
} else if fp.Protocol == iputil.IPProtocolICMP { //note that orientation doesn't matter on ICMP
fp.RemotePort = binary.BigEndian.Uint16(data[ihl+4 : ihl+6]) //identifier switch {
fp.LocalPort = 0 //code would be uint16(data[ihl+1]) case fp.Protocol == iputil.IPProtocolICMP: //note that orientation doesn't matter on ICMP
} else if incoming { fp.RemotePort = binary.BigEndian.Uint16(data[ihl+4 : ihl+6]) //identifier, LocalPort stays 0 but could be the code, data[ihl+1]
fp.RemotePort = binary.BigEndian.Uint16(data[ihl : ihl+2]) //src port
fp.LocalPort = binary.BigEndian.Uint16(data[ihl+2 : ihl+4]) //dst port case iputil.HasPorts(fp.Protocol):
} else { if incoming {
fp.LocalPort = binary.BigEndian.Uint16(data[ihl : ihl+2]) //src port fp.RemotePort = binary.BigEndian.Uint16(data[ihl : ihl+2]) //src port
fp.RemotePort = binary.BigEndian.Uint16(data[ihl+2 : ihl+4]) //dst port fp.LocalPort = binary.BigEndian.Uint16(data[ihl+2 : ihl+4]) //dst port
} else {
fp.LocalPort = binary.BigEndian.Uint16(data[ihl : ihl+2]) //src port
fp.RemotePort = binary.BigEndian.Uint16(data[ihl+2 : ihl+4]) //dst port
}
} }
return nil return nil
@@ -476,9 +473,7 @@ func parseV4(data []byte, incoming bool, fp *firewall.ParsedPacket) error {
func (f *Interface) handleOutsideMessagePacket(hostinfo *HostInfo, messageCounter uint64, out []byte, rxc *rxContext) { func (f *Interface) handleOutsideMessagePacket(hostinfo *HostInfo, messageCounter uint64, out []byte, rxc *rxContext) {
err := newPacket(out, true, rxc.fwPacket) err := newPacket(out, true, rxc.fwPacket)
if err != nil { if err != nil {
if f.l.Enabled(context.Background(), slog.LevelDebug) { hostinfo.logger(f.l).Warn("Error while validating inbound packet", "error", err, "packet", out)
hostinfo.logger(f.l).Debug("Error while validating inbound packet", "error", err, "packet", out)
}
return return
} }
+86 -21
View File
@@ -77,7 +77,7 @@ func Test_newPacket(t *testing.T) {
// account for variable ip header length - outgoing // account for variable ip header length - outgoing
h = ipv4.Header{ h = ipv4.Header{
Version: 1, Version: 1,
Protocol: 2, Protocol: iputil.IPProtocolUDP,
Len: 100, Len: 100,
Src: net.IPv4(10, 0, 0, 1), Src: net.IPv4(10, 0, 0, 1),
Dst: net.IPv4(10, 0, 0, 2), Dst: net.IPv4(10, 0, 0, 2),
@@ -89,7 +89,7 @@ func Test_newPacket(t *testing.T) {
err = newPacket(b, false, p) err = newPacket(b, false, p)
require.NoError(t, err) require.NoError(t, err)
assert.Equal(t, uint8(2), p.Protocol) assert.Equal(t, uint8(iputil.IPProtocolUDP), p.Protocol)
assert.Equal(t, netip.MustParseAddr("10.0.0.1"), p.LocalAddr) assert.Equal(t, netip.MustParseAddr("10.0.0.1"), p.LocalAddr)
assert.Equal(t, netip.MustParseAddr("10.0.0.2"), p.RemoteAddr) assert.Equal(t, netip.MustParseAddr("10.0.0.2"), p.RemoteAddr)
assert.Equal(t, uint16(6), p.RemotePort) assert.Equal(t, uint16(6), p.RemotePort)
@@ -97,6 +97,71 @@ func Test_newPacket(t *testing.T) {
assert.False(t, p.Fragment) assert.False(t, p.Fragment)
} }
// Test_newPacket_v4NoPorts ensures only protocols that lead with src/dst ports report them. The first 4 bytes of other protocols are not ports,
// and are often chosen by the sender: a GRE header's protocol type, an IPIP inner total length, a 6in4 flow label.
// Reading them as ports would let a `proto: any` rule with a port match tunneled traffic.
func Test_newPacket_v4NoPorts(t *testing.T) {
p := &firewall.ParsedPacket{}
for _, proto := range []uint8{4, 41, 47, 50, 51} { // IPIP, 6in4, GRE, ESP, AH
h := ipv4.Header{
Version: 4,
Len: ipv4.HeaderLen,
Protocol: int(proto),
Src: net.IPv4(10, 0, 0, 1),
Dst: net.IPv4(10, 0, 0, 2),
}
b, err := h.Marshal()
require.NoError(t, err)
b = append(b, []byte{0x00, 0x00, 0x86, 0xdd}...) // a GRE header carrying IPv6 would read as port 34525
for _, incoming := range []bool{true, false} {
require.NoError(t, newPacket(b, incoming, p))
assert.Equal(t, proto, p.Protocol)
assert.Equal(t, uint16(0), p.RemotePort)
assert.Equal(t, uint16(0), p.LocalPort)
assert.False(t, p.Fragment)
}
}
}
// Test_newPacket_portedProtocols ensures UDP-Lite, DCCP, and SCTP report ports from the same leading
// src/dst port words as TCP and UDP, oriented the same way.
func Test_newPacket_portedProtocols(t *testing.T) {
p := &firewall.ParsedPacket{}
ports := []byte{0x13, 0x88, 0x00, 0x50} // src 5000, dst 80
for _, proto := range []uint8{iputil.IPProtocolUDPLite, iputil.IPProtocolDCCP, iputil.IPProtocolSCTP} {
h := ipv4.Header{
Version: 4,
Len: ipv4.HeaderLen,
Protocol: int(proto),
Src: net.IPv4(10, 0, 0, 1),
Dst: net.IPv4(10, 0, 0, 2),
}
v4, err := h.Marshal()
require.NoError(t, err)
v4 = append(v4, ports...)
v6 := make([]byte, 40, 44)
v6[0] = 0x60
v6[6] = proto
v6[7] = 64
v6 = append(v6, ports...)
for _, b := range [][]byte{v4, v6} {
require.NoError(t, newPacket(b, true, p))
assert.Equal(t, proto, p.Protocol)
assert.Equal(t, uint16(5000), p.RemotePort)
assert.Equal(t, uint16(80), p.LocalPort)
require.NoError(t, newPacket(b, false, p))
assert.Equal(t, uint16(5000), p.LocalPort)
assert.Equal(t, uint16(80), p.RemotePort)
}
}
}
func Test_newPacket_v6(t *testing.T) { func Test_newPacket_v6(t *testing.T) {
p := &firewall.ParsedPacket{} p := &firewall.ParsedPacket{}
@@ -712,46 +777,46 @@ func Test_newPacket_v6ExtHeaderPastBuffer(t *testing.T) {
// Test_newPacket_v6ExtHeaderConfusion is a regression test for parseV6 walking any unrecognized // Test_newPacket_v6ExtHeaderConfusion is a regression test for parseV6 walking any unrecognized
// Next Header as if it were an ipv6 extension header. A real upper layer protocol Nebula doesn't // Next Header as if it were an ipv6 extension header. A real upper layer protocol Nebula doesn't
// dissect (SCTP here) is not walkable, so applying the (len+1)*8 formula marched into the SCTP // dissect (GRE here) is not walkable, so applying the (len+1)*8 formula marched into the GRE
// payload and landed on a byte that looked like UDP, forging a protocol/port pair the firewall // payload and landed on a byte that looked like UDP, forging a protocol/port pair the firewall
// would trust while the host delivered the real SCTP datagram. The fix fails closed: the packet // would trust while the host delivered the real GRE packet. The fix fails closed: the packet
// is classified as its true protocol with no ports, so it only matches an `any` rule. // is classified as its true protocol with no ports, so it only matches an `any` rule.
func Test_newPacket_v6ExtHeaderConfusion(t *testing.T) { func Test_newPacket_v6ExtHeaderConfusion(t *testing.T) {
p := &firewall.ParsedPacket{} p := &firewall.ParsedPacket{}
pkt := make([]byte, 52) pkt := make([]byte, 52)
pkt[0] = 0x60 // version 6 pkt[0] = 0x60 // version 6
pkt[6] = byte(layers.IPProtocolSCTP) // NextHeader = SCTP, a real protocol, not an extension header pkt[6] = byte(layers.IPProtocolGRE) // NextHeader = GRE, a real protocol, not an extension header
pkt[7] = 64 // hop limit pkt[7] = 64 // hop limit
// Real SCTP header at offset 40. Pre-fix parseV6 walked SCTP as an extension header: byte 41 (0x00, the // Real GRE header at offset 40. Pre-fix parseV6 walked GRE as an extension header: byte 41 (0x00, the
// low byte of the src port below) was read as the header length, giving next=(0+1)*8=8, which landed the // low byte of the flags/version word below) was read as the header length, giving next=(0+1)*8=8, which
// walk on byte 40 (0x11), misread as NextHeader=UDP, then bytes 48-51 as ports. // landed the walk on byte 40 (0x11), misread as NextHeader=UDP, then bytes 48-51 as ports.
binary.BigEndian.PutUint16(pkt[40:42], 0x1100) // SCTP src port; byte 40=0x11, byte 41=0x00 binary.BigEndian.PutUint16(pkt[40:42], 0x1100) // GRE flags/version; byte 40=0x11, byte 41=0x00
binary.BigEndian.PutUint16(pkt[42:44], 445) // SCTP dst port, never read by parseV6 binary.BigEndian.PutUint16(pkt[42:44], 0x86dd) // GRE protocol type, never read by parseV6
binary.BigEndian.PutUint16(pkt[48:50], 53) // SCTP checksum bytes, pre-fix forged RemotePort binary.BigEndian.PutUint16(pkt[48:50], 53) // GRE payload, pre-fix forged RemotePort
binary.BigEndian.PutUint16(pkt[50:52], 53) // pre-fix forged LocalPort binary.BigEndian.PutUint16(pkt[50:52], 53) // pre-fix forged LocalPort
require.NoError(t, newPacket(pkt, true, p)) require.NoError(t, newPacket(pkt, true, p))
assert.Equal(t, uint8(layers.IPProtocolSCTP), p.Protocol, "must classify as the true protocol, not the forged UDP") assert.Equal(t, uint8(layers.IPProtocolGRE), p.Protocol, "must classify as the true protocol, not the forged UDP")
assert.Equal(t, uint16(0), p.RemotePort) assert.Equal(t, uint16(0), p.RemotePort)
assert.Equal(t, uint16(0), p.LocalPort) assert.Equal(t, uint16(0), p.LocalPort)
assert.False(t, p.Fragment) assert.False(t, p.Fragment)
// Same confusion, but the unknown protocol sits after a real extension header. The HopByHop is walked // Same confusion, but the unknown protocol sits after a real extension header. The HopByHop is walked
// correctly, then SCTP must still fail closed instead of being walked into its own payload. Protocol is // correctly, then GRE must still fail closed instead of being walked into its own payload. Protocol is
// the only assertion that discriminates the fix here, a regression that walked SCTP would misclassify it. // the only assertion that discriminates the fix here, a regression that walked GRE would misclassify it.
chained := make([]byte, 60) chained := make([]byte, 60)
chained[0] = 0x60 // version 6 chained[0] = 0x60 // version 6
chained[6] = byte(layers.IPProtocolIPv6HopByHop) // NextHeader = HopByHop extension chained[6] = byte(layers.IPProtocolIPv6HopByHop) // NextHeader = HopByHop extension
chained[7] = 64 // hop limit chained[7] = 64 // hop limit
chained[40] = byte(layers.IPProtocolSCTP) // HopByHop NextHeader = SCTP chained[40] = byte(layers.IPProtocolGRE) // HopByHop NextHeader = GRE
chained[41] = 0 // HopByHop length 0 -> 8 bytes, SCTP begins at offset 48 chained[41] = 0 // HopByHop length 0 -> 8 bytes, GRE begins at offset 48
binary.BigEndian.PutUint16(chained[48:50], 0x1100) // SCTP src port, pre-fix forged NextHeader/length bait binary.BigEndian.PutUint16(chained[48:50], 0x1100) // GRE flags/version, pre-fix forged NextHeader/length bait
binary.BigEndian.PutUint16(chained[50:52], 445) // SCTP dst port, never read by parseV6 binary.BigEndian.PutUint16(chained[50:52], 0x86dd) // GRE protocol type, never read by parseV6
require.NoError(t, newPacket(chained, true, p)) require.NoError(t, newPacket(chained, true, p))
assert.Equal(t, uint8(layers.IPProtocolSCTP), p.Protocol, "must fail closed on the unknown protocol after the extension header") assert.Equal(t, uint8(layers.IPProtocolGRE), p.Protocol, "must fail closed on the unknown protocol after the extension header")
assert.Equal(t, uint16(0), p.RemotePort) assert.Equal(t, uint16(0), p.RemotePort)
assert.Equal(t, uint16(0), p.LocalPort) assert.Equal(t, uint16(0), p.LocalPort)
assert.False(t, p.Fragment) assert.False(t, p.Fragment)
-57
View File
@@ -1,57 +0,0 @@
//go:build !e2e_testing
package overlay
import (
"fmt"
"log/slog"
"golang.org/x/sys/windows"
"golang.org/x/sys/windows/registry"
)
var tcpipInterfaceKeys = []string{
`SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\Interfaces`,
`SYSTEM\CurrentControlSet\Services\Tcpip6\Parameters\Interfaces`,
}
// disableDNSRegistration prevents publishing overlay addresses in the host's
// DNS records, where clients outside the overlay cannot reach them.
// Failures are non-fatal so registry access cannot prevent tunnel creation.
func disableDNSRegistration(l *slog.Logger, guid windows.GUID) {
id := guid.String()
for _, base := range tcpipInterfaceKeys {
path := base + `\` + id
if err := setDNSRegistrationValues(path); err != nil {
l.Warn("Failed to disable DNS registration on the tun adapter", "error", err, "key", path)
continue
}
l.Debug("Disabled DNS registration on the tun adapter", "key", path)
}
}
func setDNSRegistrationValues(path string) error {
// The stack may not have created the adapter's interface key yet.
k, _, err := registry.CreateKey(registry.LOCAL_MACHINE, path, registry.SET_VALUE)
if err != nil {
return fmt.Errorf("open HKLM\\%s: %w", path, err)
}
defer k.Close()
// Neither RegistrationEnabled nor DisableDynamicUpdate alone reliably
// prevents registration, so all three settings are required.
values := []struct {
name string
data uint32
}{
{"RegistrationEnabled", 0},
{"DisableDynamicUpdate", 1},
{"MaxNumberOfAddressesToRegister", 0},
}
for _, v := range values {
if err := k.SetDWordValue(v.name, v.data); err != nil {
return fmt.Errorf("set %s: %w", v.name, err)
}
}
return nil
}
+4 -28
View File
@@ -50,13 +50,7 @@ func (r Route) String() string {
func makeRouteTree(l *slog.Logger, routes []Route, allowMTU bool) (*bart.Table[routing.Gateways], error) { func makeRouteTree(l *slog.Logger, routes []Route, allowMTU bool) (*bart.Table[routing.Gateways], error) {
routeTree := new(bart.Table[routing.Gateways]) routeTree := new(bart.Table[routing.Gateways])
seen := make(map[netip.Prefix]bool, len(routes))
for _, r := range routes { for _, r := range routes {
if seen[r.Cidr] {
l.Warn("route is listed more than once, only the last entry's via is used", "route", r)
}
seen[r.Cidr] = true
if !allowMTU && r.MTU > 0 { if !allowMTU && r.MTU > 0 {
l.Warn("route MTU is not supported on this platform", l.Warn("route MTU is not supported on this platform",
"goos", runtime.GOOS, "goos", runtime.GOOS,
@@ -104,11 +98,7 @@ func parseRoutes(c *config.C, networks []netip.Prefix) ([]Route, error) {
mtu, ok := rMtu.(int) mtu, ok := rMtu.(int)
if !ok { if !ok {
s, ok := rMtu.(string) mtu, err = strconv.Atoi(rMtu.(string))
if !ok {
return nil, fmt.Errorf("entry %v.mtu in tun.routes is not an integer: %v", i+1, rMtu)
}
mtu, err = strconv.Atoi(s)
if err != nil { if err != nil {
return nil, fmt.Errorf("entry %v.mtu in tun.routes is not an integer: %v", i+1, err) return nil, fmt.Errorf("entry %v.mtu in tun.routes is not an integer: %v", i+1, err)
} }
@@ -184,11 +174,7 @@ func parseUnsafeRoutes(c *config.C, networks []netip.Prefix) ([]Route, error) {
if rMtu, ok := m["mtu"]; ok { if rMtu, ok := m["mtu"]; ok {
mtu, ok = rMtu.(int) mtu, ok = rMtu.(int)
if !ok { if !ok {
s, ok := rMtu.(string) mtu, err = strconv.Atoi(rMtu.(string))
if !ok {
return nil, fmt.Errorf("entry %v.mtu in tun.unsafe_routes is not an integer: %v", i+1, rMtu)
}
mtu, err = strconv.Atoi(s)
if err != nil { if err != nil {
return nil, fmt.Errorf("entry %v.mtu in tun.unsafe_routes is not an integer: %v", i+1, err) return nil, fmt.Errorf("entry %v.mtu in tun.unsafe_routes is not an integer: %v", i+1, err)
} }
@@ -206,15 +192,10 @@ func parseUnsafeRoutes(c *config.C, networks []netip.Prefix) ([]Route, error) {
metric, ok := rMetric.(int) metric, ok := rMetric.(int)
if !ok { if !ok {
s, ok := rMetric.(string) _, err = strconv.ParseInt(rMetric.(string), 10, 32)
if !ok {
return nil, fmt.Errorf("entry %v.metric in tun.unsafe_routes is not an integer: %v", i+1, rMetric)
}
n, err := strconv.ParseInt(s, 10, 32)
if err != nil { if err != nil {
return nil, fmt.Errorf("entry %v.metric in tun.unsafe_routes is not an integer: %v", i+1, err) return nil, fmt.Errorf("entry %v.metric in tun.unsafe_routes is not an integer: %v", i+1, err)
} }
metric = int(n)
} }
if metric < 0 || metric > math.MaxInt32 { if metric < 0 || metric > math.MaxInt32 {
@@ -267,15 +248,10 @@ func parseUnsafeRoutes(c *config.C, networks []netip.Prefix) ([]Route, error) {
gatewayWeight, ok := rGatewayWeight.(int) gatewayWeight, ok := rGatewayWeight.(int)
if !ok { if !ok {
s, ok := rGatewayWeight.(string) _, err = strconv.ParseInt(rGatewayWeight.(string), 10, 32)
if !ok {
return nil, fmt.Errorf("entry .weight in tun.unsafe_routes[%v].via[%v] is not an integer", i+1, ig+1)
}
w, err := strconv.ParseInt(s, 10, 32)
if err != nil { if err != nil {
return nil, fmt.Errorf("entry .weight in tun.unsafe_routes[%v].via[%v] is not an integer", i+1, ig+1) return nil, fmt.Errorf("entry .weight in tun.unsafe_routes[%v].via[%v] is not an integer", i+1, ig+1)
} }
gatewayWeight = int(w)
} }
if gatewayWeight < 1 || gatewayWeight > math.MaxInt32 { if gatewayWeight < 1 || gatewayWeight > math.MaxInt32 {
-116
View File
@@ -1,17 +1,13 @@
package overlay package overlay
import ( import (
"bytes"
"fmt" "fmt"
"net/netip" "net/netip"
"strings"
"testing" "testing"
"github.com/slackhq/nebula/config" "github.com/slackhq/nebula/config"
"github.com/slackhq/nebula/header"
"github.com/slackhq/nebula/routing" "github.com/slackhq/nebula/routing"
"github.com/slackhq/nebula/test" "github.com/slackhq/nebula/test"
"github.com/slackhq/nebula/udp"
"github.com/stretchr/testify/assert" "github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require" "github.com/stretchr/testify/require"
) )
@@ -406,115 +402,3 @@ func Test_makeMultipathUnsafeRouteTree(t *testing.T) {
routing.CalculateBucketsForGateways(expectedGateways) routing.CalculateBucketsForGateways(expectedGateways)
assert.ElementsMatch(t, expectedGateways, r) assert.ElementsMatch(t, expectedGateways, r)
} }
func Test_getMTU(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
// unset
assert.Equal(t, DefaultMTU, getMTU(l, c))
// the largest a relayed packet can carry in one underlay buffer
require.Equal(t, udp.MTU, MaxMTU+header.MaxOverhead)
c.Settings["tun"] = map[string]any{"mtu": MaxMTU}
assert.Equal(t, MaxMTU, getMTU(l, c))
// too big is capped
c.Settings["tun"] = map[string]any{"mtu": 9000}
assert.Equal(t, MaxMTU, getMTU(l, c))
}
func Test_getAllRoutesFromConfigCapsMTU(t *testing.T) {
l := test.NewLogger()
c := config.NewC(l)
c.Settings["tun"] = map[string]any{
"routes": []any{
map[string]any{"mtu": "9000", "route": "10.0.0.0/29"},
map[string]any{"mtu": "1300", "route": "10.0.0.8/29"},
},
"unsafe_routes": []any{
map[string]any{"via": "127.0.0.1", "mtu": "9000", "route": "1.0.0.0/29"},
},
}
changed, routes, err := getAllRoutesFromConfig(l, c, []netip.Prefix{netip.MustParsePrefix("10.0.0.0/24")}, true)
require.NoError(t, err)
assert.True(t, changed)
mtus := map[string]int{}
for _, r := range routes {
mtus[r.Cidr.String()] = r.MTU
}
assert.Equal(t, map[string]int{"10.0.0.0/29": MaxMTU, "10.0.0.8/29": 1300, "1.0.0.0/29": MaxMTU}, mtus)
}
// The route tree keeps one entry per CIDR, so a CIDR listed twice only routes by the last one. Say so instead of
// letting the earlier via, or on Windows the earlier metric, quietly lose
func Test_makeRouteTree_WarnsOnDuplicate(t *testing.T) {
buf := &bytes.Buffer{}
l := test.NewLoggerWithOutput(buf)
via := func(s string) routing.Gateways {
return routing.Gateways{routing.NewGateway(netip.MustParseAddr(s), 1)}
}
cidr := netip.MustParsePrefix("10.251.1.0/24")
_, err := makeRouteTree(l, []Route{
{Cidr: cidr, Via: via("10.250.0.2"), Install: true, Metric: 100},
{Cidr: netip.MustParsePrefix("10.251.2.0/24"), Via: via("10.250.0.2"), Install: true},
}, false)
require.NoError(t, err)
assert.Empty(t, buf.String())
tree, err := makeRouteTree(l, []Route{
{Cidr: cidr, Via: via("10.250.0.2"), Install: true, Metric: 100},
{Cidr: cidr, Via: via("10.250.0.3"), Install: true, Metric: 200},
}, false)
require.NoError(t, err)
assert.Equal(t, 1, strings.Count(buf.String(), "route is listed more than once, only the last entry's via is used"))
gw, ok := tree.Lookup(netip.MustParseAddr("10.251.1.1"))
require.True(t, ok)
assert.Equal(t, netip.MustParseAddr("10.250.0.3"), gw[0].Addr(), "the last entry wins")
}
// A quoted metric or weight is a number written as a string, it means that number. Anything else is an error, not a
// panic
func Test_parseUnsafeRoutes_QuotedNumbers(t *testing.T) {
c := config.NewC(test.NewLogger())
n := netip.MustParsePrefix("10.0.0.1/24")
route := func(metric, weight any) {
c.Settings["tun"] = map[string]any{"unsafe_routes": []any{map[string]any{
"route": "1.0.0.0/8",
"metric": metric,
"via": []any{map[string]any{"gateway": "10.0.0.2", "weight": weight}},
}}}
}
route("100", "5")
routes, err := parseUnsafeRoutes(c, []netip.Prefix{n})
require.NoError(t, err)
require.Len(t, routes, 1)
assert.Equal(t, 100, routes[0].Metric)
assert.Equal(t, "{addr: 10.0.0.2, weight: 5}", routes[0].Via[0].String())
route(1.5, 1)
_, err = parseUnsafeRoutes(c, []netip.Prefix{n})
require.EqualError(t, err, "entry 1.metric in tun.unsafe_routes is not an integer: 1.5")
route(1, true)
_, err = parseUnsafeRoutes(c, []netip.Prefix{n})
require.EqualError(t, err, "entry .weight in tun.unsafe_routes[1].via[1] is not an integer")
}
// An mtu that is neither a number nor a quoted one is an error, not a panic
func Test_parseRoutes_MTUNotANumber(t *testing.T) {
c := config.NewC(test.NewLogger())
n := netip.MustParsePrefix("10.0.0.1/24")
c.Settings["tun"] = map[string]any{"routes": []any{map[string]any{"route": "1.0.0.0/8", "mtu": 1500.5}}}
_, err := parseRoutes(c, []netip.Prefix{n})
require.EqualError(t, err, "entry 1.mtu in tun.routes is not an integer: 1500.5")
c.Settings["tun"] = map[string]any{"unsafe_routes": []any{map[string]any{"route": "1.0.0.0/8", "via": "10.0.0.2", "mtu": 1500.5}}}
_, err = parseUnsafeRoutes(c, []netip.Prefix{n})
require.EqualError(t, err, "entry 1.mtu in tun.unsafe_routes is not an integer: 1500.5")
}
+3 -13
View File
@@ -20,7 +20,6 @@ type offloadQueueSet struct {
// usoEnabled is true when newTun successfully negotiated TUN_F_USO4|6 with the kernel. // usoEnabled is true when newTun successfully negotiated TUN_F_USO4|6 with the kernel.
// Queues created by Add inherit this and surface it via Offload.USOSupported so coalescers can gate USO emission. // Queues created by Add inherit this and surface it via Offload.USOSupported so coalescers can gate USO emission.
usoEnabled bool usoEnabled bool
shutdown atomic.Bool
closed atomic.Bool closed atomic.Bool
// l is handed to each queue for its bad-vnet-header drop logging. // l is handed to each queue for its bad-vnet-header drop logging.
l *slog.Logger l *slog.Logger
@@ -70,17 +69,6 @@ func (c *offloadQueueSet) wakeForShutdown() error {
return err return err
} }
// Shutdown wakes every reader without releasing any fds, they see POLLIN on the eventfd and return os.ErrClosed
func (c *offloadQueueSet) Shutdown() error {
if c.shutdown.Swap(true) {
return nil
}
for _, x := range c.pq {
x.shutdown.Store(true)
}
return c.wakeForShutdown()
}
func (c *offloadQueueSet) Close() error { func (c *offloadQueueSet) Close() error {
if c.closed.Swap(true) { if c.closed.Swap(true) {
return nil return nil
@@ -88,7 +76,9 @@ func (c *offloadQueueSet) Close() error {
errs := []error{} errs := []error{}
if err := c.Shutdown(); err != nil { // Signal all readers blocked in poll to wake up and exit.
// They observe POLLIN on the shutdown eventfd and return os.ErrClosed.
if err := c.wakeForShutdown(); err != nil {
errs = append(errs, err) errs = append(errs, err)
} }
Loaded 100 of 163 files, more files were not shown because too many files have changed in this diff. Show more