[
https://issues.apache.org/jira/browse/LIVY-866?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Gyorgy Gal updated LIVY-866:
----------------------------
Fix Version/s: 0.10.0
(was: 0.9.0)
This issue has been moved to the 0.10.0 release as part of a bulk update. If
you feel this is moved out inappropriately, feel free to provide justification
and reset the Fix Version to 0.9.0.
> Optimize Yarn GetApplications Query to prevent additional load on Yarn and
> Livy
> -------------------------------------------------------------------------------
>
> Key: LIVY-866
> URL: https://issues.apache.org/jira/browse/LIVY-866
> Project: Livy
> Issue Type: Improvement
> Components: Server
> Reporter: Akshat Bordia
> Priority: Major
> Fix For: 0.10.0
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> Currently Livy queries Yarn applications by applicationType : Spark. This
> will put heavy load on Yarn clusters if there are thousands or more Spark
> application in all states (running, finished, failed, queued etc.).
> A better approach would be to query the applications by tags in addition to
> job type since Livy only needs to track application with certain application
> tags. However, YarnClient does not expose any API to query applications by
> tags.
>
> We can extend the YarnClientImpl which would allow us to use protected
> rmClient in YarnClient and we can implement getApplications method with
> GetApplicationsRequest where we can include tags.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
