[
https://issues.apache.org/jira/browse/HBASE-24001?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17066841#comment-17066841
]
Michael Stack commented on HBASE-24001:
---------------------------------------
[~kgyrtkirk] Ok by you if I resolve this now?
All of our nightlies were changed to poll and only run if changes made
(previous they would run regardless) and the flakey suite tests has been turned
off on a bunch of branches and made less frequent on non-hot branches. I've
been studying https://builds.apache.org/label/Hadoop/load-statistics?type=min
since changes were made. Now there are always free executors; for sure there is
no back-up waiting on free executors as there was previous to the above changes.
> The "HBase Nightly" job may exhaust all executors - and create a deadlock
> -------------------------------------------------------------------------
>
> Key: HBASE-24001
> URL: https://issues.apache.org/jira/browse/HBASE-24001
> Project: HBase
> Issue Type: Bug
> Reporter: Zoltan Haindrich
> Priority: Major
>
> What seems to happen is:
> * the job needs more and more executors to run
> * in case there are several of them launched in parallel (different branches)
> say 5
> * they take up all available executors labeled "Hadoop" and then they wait
> for more indefinetly
> I've killed a few builds...see my notes here:
> https://the-asf.slack.com/archives/C7M1AFYLS/p1584442366000900
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
