[
https://issues.apache.org/jira/browse/UIMA-2767?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13609394#comment-13609394
]
Lou DeGenaro commented on UIMA-2767:
------------------------------------
OR health monitor will check for (init fails > cap) and (shares assigned by RM
== 0), then cancel job accordingly.
> DUCC resource manager (RM) and orchestrator (OR) not handling Job Process
> (JP) capping correctly
> ------------------------------------------------------------------------------------------------
>
> Key: UIMA-2767
> URL: https://issues.apache.org/jira/browse/UIMA-2767
> Project: UIMA
> Issue Type: Bug
> Reporter: Lou DeGenaro
> Assignee: Lou DeGenaro
> Priority: Minor
>
> Job #30692 initialized correctly and began processing work items. During the
> course of the Job's run JPs that initialized correctly were terminated, and
> newly allocated JPs all failed to initialize. RM was endlessly stuck
> allocating new JPs, above and beyond the init failure cap.
> Two things needs to be done. First, RM should NEVER allocate more JPs to a
> Job that has been capped. Second, OR should terminate any Job that has no
> JPs and never will have any more JPs due to capping.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
