[
https://issues.apache.org/jira/browse/YARN-4220?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14943721#comment-14943721
]
Li Lu commented on YARN-4220:
-----------------------------
Well this may be the thing I'm missing now: I made a call
{{http://localhost:8188/ws/v2/timeline/entities/application_1443660447597_0001/YARN_APPLICATION?userid=llu&fields=ALL}}
and get exceptions {{java.lang.NullPointerException: flowId shouldn't be
null}}. I believe we do have the ability to handle this kind of requests, but
for some reason we're blocking this in some intermediate levels. A call like
{{http://localhost:8188/ws/v2/timeline/entities/application_1443660447597_0001/YARN_APPLICATION?userid=llu&fields=ALL&flowid=flow_1443660447597_1&flowrunid=1}}
would return the correct info though. Am I calling it in the wrong way?
Thanks!
> [Storage implementation] Support getEntities with only Application id but no
> flow and flow run ID
> -------------------------------------------------------------------------------------------------
>
> Key: YARN-4220
> URL: https://issues.apache.org/jira/browse/YARN-4220
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: timelineserver
> Reporter: Li Lu
> Assignee: Li Lu
> Priority: Minor
>
> Currently we're enforcing flow and flowrun id to be non-null values on
> {{getEntities}}. We can actually query the appToFlow table to figure out an
> application's flow id and flowrun id if they're missing. This will simplify
> normal queries.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
