hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Abhishek Singh Chouhan (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-14718) Increase sleep time in TestRegionMover#testLoadWithoutAck
Date Fri, 30 Oct 2015 07:16:27 GMT

    [ https://issues.apache.org/jira/browse/HBASE-14718?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14982028#comment-14982028
] 

Abhishek Singh Chouhan commented on HBASE-14718:
------------------------------------------------

Apologies for failing the builds.

{noformat}
LOG.info("Successfully Unloaded\nNow Loading");
rm.load();
Thread.sleep(100);
assertEquals(noRegions, regionServer.getNumberOfOnlineRegions());
{noformat}

This was failing on slow machines as the loading was taking some time. Changed the test to
make it less strict too (since it is for best effort mode, only checking if some regions are
loaded)
{noformat}
rm.load();
    TEST_UTIL.waitFor(5000, 500, new Predicate<Exception>() {
      @Override
      public boolean evaluate() throws Exception {
        return regionServer.getNumberOfOnlineRegions() > 0;
      }
    });
{noformat}

Attaching patch in HBASE-13014.

> Increase sleep time in TestRegionMover#testLoadWithoutAck
> ---------------------------------------------------------
>
>                 Key: HBASE-14718
>                 URL: https://issues.apache.org/jira/browse/HBASE-14718
>             Project: HBase
>          Issue Type: Test
>          Components: test
>    Affects Versions: 2.0.0
>            Reporter: Samir Ahmic
>            Assignee: Abhishek Singh Chouhan
>            Priority: Critical
>             Fix For: 2.0.0
>
>         Attachments: HBASE-14718.patch
>
>
> I have notice this failed while working on [HBASE-14664| https://issues.apache.org/jira/browse/HBASE-14664].
Looks like test outcome depends on how fast is machine running this test. On my laptom threshold
is 50ms when test start to fail. I assume on Apache Jenkins 100ms is very close to treshold
based on exception:
> {code}
> java.lang.AssertionError: expected:<3> but was:<2>
> 	at org.junit.Assert.fail(Assert.java:88)
> 	at org.junit.Assert.failNotEquals(Assert.java:834)
> 	at org.junit.Assert.assertEquals(Assert.java:645)
> 	at org.junit.Assert.assertEquals(Assert.java:631)
> 	at org.apache.hadoop.hbase.util.TestRegionMover.testLoadWithoutAck(TestRegionMover.java:110)
> {code} 
> So my suggestion is to increase this to 150ms.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message