hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hong Tang (JIRA)" <j...@apache.org>
Subject [jira] Commented: (MAPREDUCE-1317) Reducing memory consumption of rumen objects
Date Fri, 18 Dec 2009 11:34:18 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-1317?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12792441#action_12792441
] 

Hong Tang commented on MAPREDUCE-1317:
--------------------------------------

Through YourKit profiling, we found two places where we could save memory:
- LoggedLocation - we should share references to the same LoggedLocation for the same preferred
location for different map tasks.
- LoggedTaskAttempt.hostName - we should keep a cache of all host names for the cluster and
share the references.

> Reducing memory consumption of rumen objects
> --------------------------------------------
>
>                 Key: MAPREDUCE-1317
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1317
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>            Reporter: Hong Tang
>
> We have encountered OutOfMemoryErrors in mumak and gridmix when dealing with very large
jobs. The purpose of this jira is to optimze memory consumption of rumen produced job objects.

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


Mime
View raw message