* try to make certificate addition/removal reloadable in some cases
* very spicy change to respond to handshakes with cert versions we cannot match with a cert that we can indeed match
* even spicier change to rehandshake if we detect our cert is lower-version than our peer, and we have a newer-version cert available
* make tryRehandshake easier to understand
* Don't delete a vpnaddr if it's part of a certificate that contains a vpnaddr that's in the static host map
* remove unused arg from ConnectionManager.shouldSwapPrimary()
* Fix for relay migration on rehandshaking issue. On rehandshake, the relay tunnel doesn't migrate to the new hostinfo object correctly, due to an incorrect Nebula IP sent in the CreateRelayRequest message.
* Add a test for this case
---------
Co-authored-by: Nate Brown <nbrown.us@gmail.com>
* cache cert verification
CheckSignature and Verify are expensive methods, and certificates are
static. Cache the results.
* use atomics
* make sure public key bytes match
* add VerifyWithCache and ResetCache
* cleanup
* use VerifyWithCache
* doc
* fix panic in handleInvalidCertificate
when HandleMonitorTick fires, the hostmap can be nil which causes a panic to occur when trying to clean up the hostmap in handleInvalidCertificate. This fix just stops the invalidation from continuing if the hostmap doesn't exist.
* removed conditional for disconnectInvalid in HandleDeletionTick
This restores the hostMap.QueryVpnIP block to how it looked before #370
was merged. I'm not sure why the patch from #370 wanted to continue on
if there was no match found in the hostmap, since there isn't anything
to do at that point (the tunnel has already been closed).
This was causing a crash because the handleInvalidCertificate check
expects the hostinfo to be passed in (but it is nil since there was no
hostinfo in the hostmap).
Fixes: #657
This change adds an index based on HostInfo.remoteIndexId. This allows
us to use HostMap.QueryReverseIndex without having to loop over all
entries in the map (this can be a bottleneck under high traffic
lighthouses).
Without this patch, a high traffic lighthouse server receiving recv_error
packets and lots of handshakes, cpu pprof trace can look like this:
flat flat% sum% cum cum%
2000ms 32.26% 32.26% 3040ms 49.03% github.com/slackhq/nebula.(*HostMap).QueryReverseIndex
870ms 14.03% 46.29% 1060ms 17.10% runtime.mapiternext
Which shows 50% of total cpu time is being spent in QueryReverseIndex.
This change adds a new helper, `(*HostInfo).logger()`, that starts a new
logrus.Entry with `vpnIp` and `certName`. We don't use the helper inside
of handshake_ix though since the certificate has not been attached to
the HostInfo yet.
Fixes: #84
