Duo Zhang created HBASE-22343:
---------------------------------
Summary: TestAMAssignWithRandExec is flaky
Key: HBASE-22343
URL: https://issues.apache.org/jira/browse/HBASE-22343
Project: HBase
Issue Type: Bug
Components: amv2, test
Reporter: Duo Zhang
I think the problem is related to HBASE-22193, where we add exponential backoff
when failing to open a region. Saw this in the output
{noformat}
2019-04-29 20:43:09,624 WARN [PEWorker-24]
assignment.TransitRegionStateProcedure(351): Failed transition, suspend 129secs
pid=15, state=RUNNABLE:REGION_STATE_TRANSITION_GET_ASSIGN_CANDIDATE,
hasLock=true;
{noformat}
And the wait time is 180 seconds, so we time out...
Maybe we should introduce a max wait time config, and set it to a small value
in this test...
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
