fix(accept): guard globalRateBucket.allow with a mutex (data race)#73
Merged
Conversation
The process-wide rate bucket's allow() is called concurrently from every Acceptor connection-handler goroutine, but tokens/lastFill were updated without synchronization — a read-modify-write data race (caught by the -race detector under concurrent load, e.g. the downstream lock-graph stress harness). Add a mutex and a concurrent regression test.
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
TeoSlayer
added a commit
to pilot-protocol/pilotprotocol
that referenced
this pull request
Jun 15, 2026
rendezvous v0.2.4 had a data race in globalRateBucket.allow() (the process-wide accept rate limiter mutated tokens/lastFill without a lock, called concurrently from every connection handler). This surfaced as the Architecture gates job's TestConcurrentDialEncryptDecrypt failing under -race. Bump to the fixed rendezvous (mutex-guarded; pilot-protocol/rendezvous#73), which pulls beacon v0.2.6 transitively. Verified: go test -race TestConcurrentDialEncryptDecrypt ./tests passes (104s, no race). Co-authored-by: Teodor Calin <teodor@vulturelabs.io>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
globalRateBucket.allow()(the process-wide accept rate limiter) mutatestokens/lastFillwithout synchronization, but it is called concurrentlyfrom every
Acceptorconnection-handler goroutine (handleJSONConn). Thatis a read-modify-write data race on the token-refill state.
Under concurrent load the
-racedetector flags it reliably — e.g. thedownstream
pilotprotocollock-graph stress harness(
TestConcurrentDialEncryptDecrypt) fails itsArchitecture gatesjobpurely because of this race in the rendezvous dependency.
Fix
sync.MutextoglobalRateBucketand lock around therefill/consume in
allow().TestGlobalRateBucket_ConcurrentAllow, a concurrent regression test(64 goroutines × 1000 calls) that fails under
-racewithout the lockand passes with it.
Testing
go test -race ./accept/green.reproduces
WARNING: DATA RACE; restoring it passes.