[
https://issues.apache.org/jira/browse/HDFS-14599?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17081742#comment-17081742
]
Xiaoqiao He commented on HDFS-14599:
------------------------------------
Hi [~weichiu],[~elgoiri]. TestDiskBalancer is still broken by HDFS-12487 in
branch-3.1 and branch-3.2, HDFS-15272 is another reported case. I suggest we
should commit this one to 3.2 and 3.1. FYI.
> HDFS-12487 breaks test
> TestDiskBalancer.testDiskBalancerWithFedClusterWithOneNameServiceEmpty
> ---------------------------------------------------------------------------------------------
>
> Key: HDFS-14599
> URL: https://issues.apache.org/jira/browse/HDFS-14599
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: diskbalancer
> Affects Versions: 3.3.0, 3.2.1, 3.1.3
> Reporter: Wei-Chiu Chuang
> Assignee: Xiaoqiao He
> Priority: Major
> Labels: newbie
> Fix For: 3.3.0
>
> Attachments: HDFS-14599.001.patch, HDFS-14599.002.patch
>
>
> It looks like HDFS-12487 changes the error message expected by
> {{TestDiskBalancer#testDiskBalancerWithFedClusterWithOneNameServiceEmpty}}.
> The test expects error "There are no blocks in the blockPool" but after
> HDFS-12487, it returns error string "NextBlock call returned null.No valid
> block to copy."
> Probably the simplest approach to fix it is to update the expected error
> string.
> Thoughts? [~bharatviswa] you crafted the test in HDFS-13715. Should we update
> the expected error string, or revert HDFS-12487?
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
