[
https://issues.apache.org/jira/browse/YARN-6111?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16016512#comment-16016512
]
Yufei Gu commented on YARN-6111:
--------------------------------
Mostly like {{JobTraceReader}} changed its way to read the trace files. It
reads json objects one by one instead of reading them as an array. So my patch
v1 removes the array mark of all jobs json object, and it works. Besides, the
example rumen file is supposed to container two jobs instead three jobs
indicated by its name. I removed the last invalid job configuration as well.
Change from:
{code}
[{job1}, {job2}, {job3}]
{code}
{code}
{job1}
{job2}
{code}
> Rumen input does't work in SLS
> ------------------------------
>
> Key: YARN-6111
> URL: https://issues.apache.org/jira/browse/YARN-6111
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: scheduler-load-simulator
> Affects Versions: 2.6.0, 2.7.3, 3.0.0-alpha2
> Environment: ubuntu14.0.4 os
> Reporter: YuJie Huang
> Assignee: Yufei Gu
> Labels: test
> Attachments: YARN-6111.001.patch
>
>
> Hi guys,
> I am trying to learn the use of SLS.
> I would like to get the file realtimetrack.json, but this it only
> contains "[]" at the end of a simulation. This is the command I use to
> run the instance:
> HADOOP_HOME $ bin/slsrun.sh --input-rumen=sample-data/2jobsmin-rumen-jh.json
> --output-dir=sample-data
> All other files, including metrics, appears to be properly populated.I can
> also trace with web:http://localhost:10001/simulate
> Can someone help?
> Thanks
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
