[
https://issues.apache.org/jira/browse/HDFS-7364?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14200425#comment-14200425
]
Mit Desai commented on HDFS-7364:
---------------------------------
Nice catch. Here balancer exits here after 5 iterations of what it thinks has
0B move. It means it is still balancing and exits in middle the process. I see
that the Bytes left to move is going down in every iteration.
It will be nice to have this fixed. But it would be good to have a unit test as
well.
> Balancer always show zero Bytes Already Moved
> ---------------------------------------------
>
> Key: HDFS-7364
> URL: https://issues.apache.org/jira/browse/HDFS-7364
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: balancer & mover
> Reporter: Tsz Wo Nicholas Sze
> Assignee: Tsz Wo Nicholas Sze
> Attachments: h7364_20141105.patch
>
>
> Here is an example:
> {noformat}
> Time Stamp Iteration# Bytes Already Moved Bytes Left To Move
> Bytes Being Moved
> Nov 5, 2014 5:23:38 PM 0 0 B 116.82 MB
> 181.07 MB
> Nov 5, 2014 5:24:30 PM 1 0 B 88.05 MB
> 181.07 MB
> Nov 5, 2014 5:25:10 PM 2 0 B 73.08 MB
> 181.07 MB
> Nov 5, 2014 5:25:49 PM 3 0 B 13.37 MB
> 90.53 MB
> Nov 5, 2014 5:26:30 PM 4 0 B 13.59 MB
> 90.53 MB
> Nov 5, 2014 5:27:12 PM 5 0 B 9.25 MB
> 90.53 MB
> The cluster is balanced. Exiting...
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
