[
https://issues.apache.org/jira/browse/HBASE-30249?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18090545#comment-18090545
]
Duo Zhang commented on HBASE-30249:
-----------------------------------
This is log sniff
{noformat}
2026-06-20T22:54:07,918 INFO [PEWorker-3]
replication.TransitPeerSyncReplicationStateProcedure(135): Successfully transit
current cluster state from STANDBY to DOWNGRADE_ACTIVE for sync replication
peer 1
2026-06-20T22:54:07,920 INFO [PEWorker-3] procedure2.ProcedureExecutor(1521):
Finished pid=24, state=SUCCESS, hasLock=false;
org.apache.hadoop.hbase.master.replication.TransitPeerSyncReplicationStateProcedure
in 727 msec
2026-06-20T22:54:09,190 INFO [Thread-898] master.HMaster(3398): ***** STOPPING
master '9c7f818fcd69,40897,1781996036525' *****
2026-06-20T22:54:09,190 INFO [Thread-898] master.HMaster(3400): STOPPED: Stop
master for test
{noformat}
Obviously, we stop master after the TransitPeerSyncReplicationStateProcedure
finishes...
> Make TestSyncReplicationStandbyKillMaster stable
> ------------------------------------------------
>
> Key: HBASE-30249
> URL: https://issues.apache.org/jira/browse/HBASE-30249
> Project: HBase
> Issue Type: Sub-task
> Components: Replication, test
> Reporter: Duo Zhang
> Assignee: Duo Zhang
> Priority: Major
>
> After analyzing, the problem is that, we want to test master switch when
> transiting, but actually, the procedure can be finished within 1 second but
> we will kill mastr after 2 seconds, so we just test nothing, and the master
> killing operation will mess up the code in tearDown and fail the test...
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
