When tun.pin_threads is on and tun.cpu_affinity is unset, pick pin CPUs
from the allowed set that do not service any up physical NIC's MSI vectors
(read-only walk of /sys/class/net/*/device/msi_irqs and
/proc/irq/*/effective_affinity_list). The old allowed[i] default pinned
encrypt threads onto exactly the cores drivers affine their first RX queue
IRQs to; a flow whose RSS queue fired there had NAPI fighting encrypt for
the core (measured 8.4 vs 10.2 Gbps REV bimodality). Falls back to the old
spread, with a log, when there aren't enough IRQ-free CPUs - e.g. drivers
that allocate one queue per core until the admin narrows them (ethtool -L).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* populate default Build version if missing
Use the Go module information built into the binary if the Build var
wasn't set during the build.
This means if you install via a specific tag, you get:
go install github.com/slackhq/nebula/cmd/nebula@v1.9.5
$ nebula -version
Version: 1.9.5
And if you install master, you get:
go install github.com/slackhq/nebula/cmd/nebula@master
$ nebula -version
Version: 1.9.5-0.20250408154034-18279ed17b10
* also default in the library
* cleanup
* upgrade to yaml.v3
The main nice fix here is that maps unmarshal into `map[string]any`
instead of `map[any]any`, so it cleans things up a bit.
* add config.AsBool
Since yaml.v3 doesn't automatically convert yes to bool now, for
backwards compat
* use type aliases for m
* more cleanup
* more cleanup
* more cleanup
* go mod cleanup
This used to work correctly because when the multiple routines work was
first added in #382, but an important part to discover the listen port
before opening the other listeners on the same socket was lost in this
PR: #653.
This change should fix the regression and allow multiple routines to
work correctly when listen.port is set to `0`.
Thanks to @rawdigits for tracking down and discovering this regression.
* Added firewall.rules.hash metric
Added a FNV-1 hash of the firewall rules as a Prometheus value.
* Switch FNV has to int64, include both hashes in log messages
* Use a uint32 for the FNV hash
Let go-metrics cast the uint32 to a int64, so it won't be lossy
when it eventually emits a float64 Prometheus metric.
By default, Nebula replies to packets it has no tunnel for with a `recv_error` packet. This packet helps speed up re-connection
in the case that Nebula on either side did not shut down cleanly. This response can be abused as a way to discover if Nebula is running
on a host though. This option lets you configure if you want to send `recv_error` packets always, never, or only to private network remotes.
valid values: always, never, private
This setting is reloadable with SIGHUP.
When running in `-test` mode, `tun` is set to nil. So we should move the
defer into the `!configTest` if block.
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x28 pc=0x54855c]
goroutine 1 [running]:
github.com/slackhq/nebula.Main.func3(0x4000135e80, {0x0, 0x0})
github.com/slackhq/nebula/main.go:176 +0x2c
github.com/slackhq/nebula.Main(0x400022e060, 0x1, {0x76faa0, 0x5}, 0x4000230000, 0x0)
github.com/slackhq/nebula/main.go:316 +0x2414
main.main()
github.com/slackhq/nebula/cmd/nebula/main.go:54 +0x540
JackDoan