RATIS-2547. Advance repliedIndex before completing write replies#1474
Merged
Conversation
3 tasks
szetszwo
approved these changes
May 29, 2026
Contributor
szetszwo
left a comment
szetszwo
left a comment
There was a problem hiding this comment.
+1 the change looks good.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
RATIS-2547
What changes were proposed in this pull request?
When
raft.server.read.optionusesREPLIED_INDEX,ReplyFlusherholds write reply callbacks and advances the leader replied index in batches. The old flow invoked reply callbacks while computing the max flushed log index, so a client could receive a successful write reply beforerepliedIndexcovered that write.This PR stores the log index separately from the reply callback, advances
repliedIndexfirst, and only then completes the held write replies. It also makesLeaderTracer.removePendingRequesttolerate a null pending request for cached-reply paths.Why are the changes needed?
A read-after-write workload can issue a linearizable read immediately after receiving a successful write reply. With
REPLIED_INDEX, that read depends on the leader replied index. Completing the write reply before advancingrepliedIndexcreates a race where the read can wait unnecessarily or time out even though the write reply was already released.How was this patch tested?
mvn -pl ratis-server,ratis-test -am -Dtest=TestReplyFlusher,TestLinearizableReadRepliedIndexWithGrpc test