[
https://issues.apache.org/jira/browse/MAPREDUCE-6424?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15265271#comment-15265271
]
Naganarasimha G R commented on MAPREDUCE-6424:
----------------------------------------------
Thanks [~varun_saxena], for the analysis but anyway i too had narrowed down to
it.
bq. It would however be better to change it to guard against any future
changes in FS based impl. But frankly FS based impl is unlikely to change.
Even i too feel the same as to read each line as entity from the file,
currently i am using
{{FileSystemTimelineReaderImpl.getTimelineRecordFromJSON}}. Test case was
failing because i had assumed that there would no other entities without any
events other than the one published by the MR job, but if the method returns
once we have found the metrics i think it should be sufficient enough. thoughts?
Uploaded the patch for the same, if req can modify the test case to use
{{FileSystemTimelineReaderImpl}}.
> Store MR counters as timeline metrics instead of event
> ------------------------------------------------------
>
> Key: MAPREDUCE-6424
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-6424
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Reporter: Junping Du
> Assignee: Naganarasimha G R
> Labels: yarn-2928-1st-milestone
> Attachments: MAPREDUCE-6424-YARN-2928.v1.001.patch,
> MAPREDUCE-6424-YARN-2928.v1.002.patch, MAPREDUCE-6424-YARN-2928.v1.003.patch,
> MAPREDUCE-6424-YARN-2928.v1.004.patch
>
>
> In MAPREDUCE-6327, we make map/reduce counters get encoded from
> JobFinishedEvent as timeline events with counters details in JSON format.
> We need to store framework specific counters as metrics in timeline service
> to support query, aggregation, etc.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org
