hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sreekanth Ramakrishnan (JIRA)" <j...@apache.org>
Subject [jira] Commented: (MAPREDUCE-1087) Limit resource usage of Map/Reduce debug script
Date Tue, 13 Oct 2009 06:16:31 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-1087?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12764967#action_12764967

Sreekanth Ramakrishnan commented on MAPREDUCE-1087:

It would also be a good idea to time out a debug script after some specific interval. A script
can go rouge and block the slot forever. Maybe something similar to what we do respect to
tasks when they fail to report status after some period of time.

> Limit resource usage of Map/Reduce debug script
> -----------------------------------------------
>                 Key: MAPREDUCE-1087
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1087
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: tasktracker
>            Reporter: Vinod K V
> As mentioned here by Devaraj [here|https://issues.apache.org/jira/browse/MAPREDUCE-915?focusedCommentId=12763756&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#action_12763756],
Map/Reduce debug scripts will run user code and hence need to be resource-limited via ulimits
and memory-monitored similar to task-memory-monitoring.

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

View raw message