satishd commented on code in PR #14004:
URL: https://github.com/apache/kafka/pull/14004#discussion_r1268911909
##########
storage/src/main/java/org/apache/kafka/server/log/remote/metadata/storage/RemoteLogLeaderEpochState.java:
##########
@@ -100,17 +100,29 @@ void
handleSegmentWithCopySegmentStartedState(RemoteLogSegmentId remoteLogSegmen
void handleSegmentWithCopySegmentFinishedState(Long startOffset,
RemoteLogSegmentId remoteLogSegmentId,
Long leaderEpochEndOffset) {
+ // If there are duplicate segments uploaded due to leader-election,
then mark them as unreferenced.
+ // Duplicate segments can be uploaded when the previous leader had
tier-lags and the next leader uploads the
+ // segment for the same leader-epoch which is a super-set of
previously uploaded segments.
+ // (eg)
+ // case-1: Duplicate segment
+ // L0 uploaded segment S0 with offsets 0-100 and L1 uploaded
segment S1 with offsets 0-200.
+ // We will mark the segment S0 as duplicate and add it to
unreferencedSegmentIds.
+ // case-2: Overlapping segments
+ // L0 uploaded segment S0 with offsets 10-90 and L1 uploaded
segment S1 with offsets 5-100, S2-101-200,
+ // and so on. When the consumer request for segment with offset
95, it should get the segment S1 and not S0.
+ Map.Entry<Long, RemoteLogSegmentId> lastEntry = offsetToId.lastEntry();
+ while (lastEntry != null && lastEntry.getKey() >= startOffset &&
highestLogOffset <= leaderEpochEndOffset) {
Review Comment:
Thanks @kamalcph for the detailed explanation. It is better to return the
epoch along with highest-offset and then check the minimum of the respective
end offset in the local leader epoch chain for that epoch and the highest
offset uploaded to remote storage for that epoch.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: jira-unsubscr...@kafka.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
