[GitHub] [hudi] xicm commented on a diff in pull request #8057: [HUDI-5831] Address flakiness of early conflict detection tests

["xicm (via GitHub)" <gi...@apache.org>]

xicm commented on code in PR #8057:
URL: https://github.com/apache/hudi/pull/8057#discussion_r1124142642


##########
hudi-timeline-service/src/main/java/org/apache/hudi/timeline/service/handlers/marker/MarkerBasedEarlyConflictDetectionRunnable.java:
##########
@@ -109,7 +109,7 @@ public void run() {
       currentFileIDs.retainAll(tableFilesIDs);
       if (!currentFileIDs.isEmpty()
           || (checkCommitConflict && MarkerUtils.hasCommitConflict(activeTimeline,
-          currentInstantAllMarkers.stream().map(MarkerUtils::makerToPartitionAndFileID).collect(Collectors.toSet()), completedCommits))) {
+          currentInstantAllMarkers.stream().map(MarkerUtils::makerToFileID).collect(Collectors.toSet()), completedCommits))) {

Review Comment:
   It's hard to create a scenario described in rfc-56, `Writer1 starts writing data at time t1 and finishes at time t3. Writer2 starts writing at time t2, and at time t4 - writer2 tries to create marker for a file fileA which already updated by writer1 to fileAa(updated)`.
   
   But to be consistent with  MarkerUtils.hasCommitConflict, the 2nd param should be a set of fileId.



-- 
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: commits-unsubscribe@hudi.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org