[
https://issues.apache.org/jira/browse/HDFS-4992?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13709415#comment-13709415
]
Tsz Wo (Nicholas), SZE commented on HDFS-4992:
----------------------------------------------
Hi Max, thanks for submitting a patch. Some minor comments:
- Could you add the new conf properties and the default values to
DFSConfigKeys? You may put them next to DFS_BALANCER_MOVEDWINWIDTH_KEY.
- Let's keep moverExecutor and dispatcherExecutor as final.
Patch looks good other than that.
> Make balancer's thread count configurable
> -----------------------------------------
>
> Key: HDFS-4992
> URL: https://issues.apache.org/jira/browse/HDFS-4992
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: balancer
> Reporter: Max Lapan
> Attachments: HDFS-4992-v1.patch
>
>
> By default, balancer has 1000 threads which moves blocks around (mover
> threads) and 200 threads which decide which block to move (dispatcher
> threads).
> On large clusters, 1000 threads creates significant load on NN, which slows
> down other HDFS activity. For example, on our cluster, 'hdfs dfs -ls /'
> command took about 1 minute when balancer is active. When no balancing in
> progress, the same command finishes in second or two.
> This patch makes amount of threads configurable by two new options
> 'dfs.balancer.moverThreads' and 'dfs.balancer.dispatcherThreads'.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
