hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Arun C Murthy (JIRA)" <j...@apache.org>
Subject [jira] Updated: (MAPREDUCE-1683) Remove JNI calls from ClusterStatus cstr
Date Wed, 07 Apr 2010 08:45:33 GMT

     [ https://issues.apache.org/jira/browse/MAPREDUCE-1683?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Arun C Murthy updated MAPREDUCE-1683:

    Attachment: MAPREDUCE-1683_yhadoop_20_9.patch

Rather than remove these fields from ClusterStatus, I've made a slightly less invasive change:
I've made the memory-related info a part of 'detailed' information alongwith tracker-names
and such. Thus users can call JobClient.getClusterStatus(true) to get this info. 

Nevertheless, the issue is important enough to make an incompatible change for better performance
on the JT. This is called from various critical pieces of framework code (e.g. JobInProgress)
to matter.

> Remove JNI calls from ClusterStatus cstr
> ----------------------------------------
>                 Key: MAPREDUCE-1683
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1683
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: jobtracker
>    Affects Versions: 0.20.2
>            Reporter: Chris Douglas
>         Attachments: MAPREDUCE-1683_yhadoop_20_9.patch
> The {{ClusterStatus}} constructor makes two JNI calls to the {{Runtime}} to fetch memory
information. {{ClusterStatus}} instances are often created inside the {{JobTracker}} to obtain
other, unrelated metrics (sometimes from schedulers' inner loops). Given that this information
is related to the {{JobTracker}} process and not the cluster, the metrics are also available
via {{JvmMetrics}}, and the jsps can gather this information for themselves: these fields
can be removed from {{ClusterStatus}}

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message