Guard per-host semaphore pruning against unbalanced releases#2283
Open
hyperxpro wants to merge 1 commit into
Open
Guard per-host semaphore pruning against unbalanced releases#2283hyperxpro wants to merge 1 commit into
hyperxpro wants to merge 1 commit into
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Motivation:
#2277 prevented the per-host connection map from growing indefinitely by reference counting entries and pruning them when they become idle. This relies on every release matching a prior reservation. An unmatched or double release can prematurely remove a live entry, allowing a new semaphore to be created and silently bypassing the per-host connection limit. The previous lookup method was also left unused in production and, if called, could create entries that were never eligible for pruning. Existing tests covered only a single host.
Modification:
Log unmatched releases so double releases are visible instead of silently corrupting the reference count. Remove the unused lookup method and update its remaining test. Add concurrent multi-host stress tests for both semaphore implementations, verifying that the per-host limit is enforced under contention and that all entries are pruned once released.
Result:
Unbalanced releases are now detectable, the unused code path has been removed, and the per-host connection limit is verified under concurrent multi-host workloads.