Mailing-List: contact hdfs-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: hdfs-issues@hadoop.apache.org
Date: Wed, 14 May 2014 15:59:15 +0000 (UTC)
From: "Junping Du (JIRA)" <jira@apache.org>
To: hdfs-issues@hadoop.apache.org
Message-ID: <JIRA.12708820.1397664872260.346089.1400083155775@arcas>
In-Reply-To: <JIRA.12708820.1397664872260@arcas>
References: <JIRA.12708820.1397664872260@arcas>
Subject: [jira] [Commented] (HDFS-6250)
 TestBalancerWithNodeGroup.testBalancerWithRackLocality fails
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HDFS-6250?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13997684#comment-13997684 ] 

Junping Du commented on HDFS-6250:
----------------------------------

Thanks for review and comments, [~airbots]!
The good case you proposed above, if it is current behavior, will be seen as a bug for balancer. As the prerequisite of balancer's handling block movement is not hurt the data reliability. It shouldn't move replicas of balancer.id across rack0 and rack1 as it will make replica's rack number reduce to 1 which affect block's reliability and inconsistent with replica placement policy. Actually, we have code below to get rid of this case:
{code}
   ...
   * 3. doing the move does not reduce the number of racks that the block has
   */
  private boolean isGoodBlockCandidate(Source source, 
{code}
Would you double check the behavior of balancer in the case you described above? If so, we should file a separated JIRA to fix Balancer. What do you think?

> TestBalancerWithNodeGroup.testBalancerWithRackLocality fails
> ------------------------------------------------------------
>
>                 Key: HDFS-6250
>                 URL: https://issues.apache.org/jira/browse/HDFS-6250
>             Project: Hadoop HDFS
>          Issue Type: Bug
>            Reporter: Kihwal Lee
>            Assignee: Chen He
>         Attachments: HDFS-6250-v2.patch, HDFS-6250-v3.patch, HDFS-6250.patch, test_log.txt
>
>
> It was seen in https://builds.apache.org/job/PreCommit-HDFS-Build/6669/
> {panel}
> java.lang.AssertionError: expected:<1800> but was:<1810>
> 	at org.junit.Assert.fail(Assert.java:93)
> 	at org.junit.Assert.failNotEquals(Assert.java:647)
> 	at org.junit.Assert.assertEquals(Assert.java:128)
> 	at org.junit.Assert.assertEquals(Assert.java:147)
> 	at org.apache.hadoop.hdfs.server.balancer.TestBalancerWithNodeGroup
>  .testBalancerWithRackLocality(TestBalancerWithNodeGroup.java:253)
> {panel}


--
This message was sent by Atlassian JIRA
(v6.2#6252)