Alexey Serbin has posted comments on this change. ( http://gerrit.cloudera.org:8080/22133 )
Change subject: KUDU-3571: fix flakiness in AutoIncrementingItest.BootstrapNoWalsNoData ...................................................................... Patch Set 1: (2 comments) http://gerrit.cloudera.org:8080/#/c/22133/1//COMMIT_MSG Commit Message: http://gerrit.cloudera.org:8080/#/c/22133/1//COMMIT_MSG@10 PS1, Line 10: unhealthy or lagging replicas I'm curious: what 'deterministic' means in this context? Write operations are being replicated regardless of regular/periodic Raft heartbeating -- the leader replica sends Raft updates right away during the replication phase, it doesn't wait for next periodic Raft heartbeat interval, no. If introducing a delay, it might be anything here, IIUC. Setting it to the Raft heartbeat interval just adds extra confusion. And I don't think this current approach in PS5 or PS6 is going to address the flakiness -- it might still pop up under some extreme conditions. If going the route of waiting for all the replicas to converge and doing it reliably, one might take a look at various methods in test scaffolding like WaitForServersToAgree() and AssertAllReplicasAgree() in What's the problem with the original approach that Yifan used in PS1? > Limiting to only leader replica might be concerning Ashwani, Abhishek -- could you add more details on your concerns around this? Thanks! http://gerrit.cloudera.org:8080/#/c/22133/6/src/kudu/integration-tests/auto_incrementing-itest.cc File src/kudu/integration-tests/auto_incrementing-itest.cc: http://gerrit.cloudera.org:8080/#/c/22133/6/src/kudu/integration-tests/auto_incrementing-itest.cc@446 PS6, Line 446: // Restart the cluster. : cluster_->Shutdown( The replication of writes happens regardless of periodic Raft heartbeating. Also, even with this extra delay there isn't a guarantee to have everything replicated across all the existing replicas -- this extra delay might provide some sense of security, but I'm not sure it's able to address the issue underlying issue. -- To view, visit http://gerrit.cloudera.org:8080/22133 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: kudu Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I5bd387c82b632dbb77aa5a45f831273392ae05b4 Gerrit-Change-Number: 22133 Gerrit-PatchSet: 1 Gerrit-Owner: Yifan Zhang <[email protected]> Gerrit-Reviewer: Abhishek Chennaka <[email protected]> Gerrit-Reviewer: Alexey Serbin <[email protected]> Gerrit-Reviewer: Ashwani Raina <[email protected]> Gerrit-Reviewer: Kudu Jenkins (120) Gerrit-Reviewer: Yifan Zhang <[email protected]> Gerrit-Comment-Date: Wed, 04 Dec 2024 06:32:08 +0000 Gerrit-HasComments: Yes
