You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@uima.apache.org by "Jerry Cwiklik (Jira)" <de...@uima.apache.org> on 2019/08/28 20:23:00 UTC

[jira] [Reopened] (UIMA-6082) UIMA-DUCC: tag process as FAILED if agent unable to determine scheduling class

     [ https://issues.apache.org/jira/browse/UIMA-6082?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jerry Cwiklik reopened UIMA-6082:
---------------------------------

Backing out the change made previously to support detection of invalid scheduling class. The RM validates scheduling class and only processes with correct class assignment are scheduled to run on a given node. 

> UIMA-DUCC: tag process as FAILED if agent unable to determine scheduling class
> ------------------------------------------------------------------------------
>
>                 Key: UIMA-6082
>                 URL: https://issues.apache.org/jira/browse/UIMA-6082
>             Project: UIMA
>          Issue Type: Improvement
>          Components: DUCC
>            Reporter: Jerry Cwiklik
>            Assignee: Jerry Cwiklik
>            Priority: Major
>             Fix For: 3.0.1-Ducc
>
>
> When an agent fails to determine scheduling class for a process as part of the OR state reconciliation, it logs a stack trace and continues. Since OR publishes its process table, the same exception is being logged and the job (service, AP) are not tagged as FAILED.
> Agent should tag a job/service/AP as FAILED in this case so that the OR does not try to allocate process(es) over and over again.



--
This message was sent by Atlassian Jira
(v8.3.2#803003)