[
https://issues.apache.org/jira/browse/HDFS-4240?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13508494#comment-13508494
]
Junping Du commented on HDFS-4240:
----------------------------------
Thanks Jing for the review. Yes. Set checking stale nodes in this class is not
necessary in this test class now and we can enable checking later if we want to
add more test cases in future. Above unit test failure is not related and I
double check this test case passed in local environment. I upload new patch
(HDFS-4240.patch) which should kick off new pre-commit tests now.
> In nodegroup-aware case, make sure nodes are avoided to place replica if some
> replica are already under the same nodegroup
> --------------------------------------------------------------------------------------------------------------------------
>
> Key: HDFS-4240
> URL: https://issues.apache.org/jira/browse/HDFS-4240
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: namenode
> Reporter: Junping Du
> Assignee: Junping Du
> Attachments: HADOOP-9045.patch, HADOOP-9045-v2.patch,
> HADOOP-9045-v3.patch, HADOOP-9045-v4.patch, HDFS-4240.patch
>
>
> In previous implementation for HADOOP-8468, 3rd replica is avoid to place on
> the same nodegroup of 2nd replica. But it didn't provide check on nodegroup
> of 1st replica, so if 2nd replica's rack is not efficient to place replica,
> then it is possible to place 3rd and 1st replica within the same node group.
> We need a change to remove all nodes from available nodes for placing replica
> if there already replica on the same nodegroup.
--
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
