[jira] [Commented] (MESOS-2787) The exception "KeyError: 'mem_rss_bytes'" is being raised when run command mesos ps --master=<>

["weitao zhou (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MESOS-2787?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14567211#comment-14567211 ] 

weitao zhou commented on MESOS-2787:
------------------------------------

patch: https://reviews.apache.org/r/34886/

> The exception "KeyError: 'mem_rss_bytes'" is being raised when run command mesos ps --master=<> 
> ------------------------------------------------------------------------------------------------
>
>                 Key: MESOS-2787
>                 URL: https://issues.apache.org/jira/browse/MESOS-2787
>             Project: Mesos
>          Issue Type: Bug
>          Components: cli, python api, slave, statistics, stout
>    Affects Versions: 0.21.1, 0.22.0, 0.22.1
>         Environment: ubuntu 14.04, mesos 0.21.1
>            Reporter: weitao zhou
>              Labels: easyfix, patch
>
> When I run the command *mesos ps --master=$MESOS-MASTER:5050*, the following exception raised:
> {code: title=exception.sh}
> File "/usr/bin/mesos-ps", line 232, in <module>
>     main()
>   File "/usr/bin/mesos-ps", line 217, in main
>     sys.stdout.write(columns[4].truncate(mem(task, statistics)))
>   File "/usr/bin/mesos-ps", line 93, in mem
>     mem_rss_bytes = entry['statistics']['mem_rss_bytes']
> KeyError: 'mem_rss_bytes'
> {code}
> I tracked it and found that some mesos slave nodes can't return the statistics yet now. In my special case, the disk volumn of some my mesos slaves is 100% full.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)