Rohith Sharma K S commented on YARN-5743:

bq. By the way we should probably publish log aggregation status and 
diagnostics too. This will help, if user cannot find the application logs in 
designated location due to some failure during aggregation.
I think we should NOT store log aggregation status and its diagnostic 
information because of couple of reasons
# aggregated log will be deleted after some time as per configuration 
configured. Since we are storing very long history of applications, very old 
applications logs might be deleted which gives wrong information to reader. 
# Diagnostic report is for per node basis. This will over spam the application 
storage. Let say, if aggregation fails 100 nodes may be because of ticket, then 
full of application entity will be filled with list of nodeId vs diagnostic 

> [Atsv2] Publish queue name and RMAppMetrics to ATS
> --------------------------------------------------
>                 Key: YARN-5743
>                 URL: https://issues.apache.org/jira/browse/YARN-5743
>             Project: Hadoop YARN
>          Issue Type: Bug
>            Reporter: Rohith Sharma K S
>            Assignee: Rohith Sharma K S
>         Attachments: 0001-YARN-5743.patch
> App queue name is missed to publish to ATSv2. 
> And RMAppMetrcs publish only cpu and memory. There are many more things to 
> publish from app metrics such as 
>  resourcePreempted;
>  numNonAMContainersPreempted;
> numAMContainersPreempted.
> And RMAppMetrics need to be published to App metrics rather than info. 

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org

Reply via email to