fix(go): tear down only the connection a request failed on - #3986
Open
hubcio wants to merge 1 commit into
Open
Conversation
Two requests failing at once are one reconnect, but the teardown that precedes it went by transport state alone. Connect marks the client connected before it signs in, so the second caller -- parked on c.mtx for the length of that sign-in -- woke to a healthy-looking state and closed the socket the first caller's reconnect had just established. That left the first caller replaying over nothing, failing it with ErrNotConnected, and started a second attempt for a reconnect that had already happened. Stamp every installed connection with a generation and carry it out of the attempt that ran on it, so the teardown is a no-op once the connection has been replaced. The leader redirect took the same stale teardown and now drops a redirect decided on a connection that is gone. Surfaced as a flaky TestConnect_ConcurrentReconnectsThroughExchangeShareOneAttempt.
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #3986 +/- ##
=========================================
Coverage 84.26% 84.26%
Complexity 1398 1398
=========================================
Files 1219 1219
Lines 173644 173670 +26
Branches 140742 140742
=========================================
+ Hits 146317 146340 +23
- Misses 23369 23371 +2
- Partials 3958 3959 +1
🚀 New features to boost your workflow:
|
numinnex
approved these changes
Aug 28, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Two requests failing at once are one reconnect, but the
teardown that precedes it went by transport state alone.
Connect marks the client connected before it signs in, so
the second caller -- parked on c.mtx for the length of that
sign-in -- woke to a healthy-looking state and closed the
socket the first caller's reconnect had just established.
That left the first caller replaying over nothing, failing
it with ErrNotConnected, and started a second attempt for a
reconnect that had already happened.
Stamp every installed connection with a generation and carry
it out of the attempt that ran on it, so the teardown is a
no-op once the connection has been replaced. The leader
redirect took the same stale teardown and now drops a
redirect decided on a connection that is gone.
Surfaced as a flaky
TestConnect_ConcurrentReconnectsThroughExchangeShareOneAttempt.