You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@kudu.apache.org by "Jean-Daniel Cryans (JIRA)" <ji...@apache.org> on 2016/07/14 18:55:20 UTC
[jira] [Comment Edited] (KUDU-1531)
org.kududb.client.ITScannerMultiTablet is flaky
[ https://issues.apache.org/jira/browse/KUDU-1531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15378126#comment-15378126 ]
Jean-Daniel Cryans edited comment on KUDU-1531 at 7/14/16 6:54 PM:
-------------------------------------------------------------------
This is worrying. We have a hash-partitioned table into 3 tablets, with a total of 20k rows, and we do a sorted scan. We "detect" going across tablet boundaries by simply checking if we're seeing a row key go down: https://github.com/cloudera/kudu/blob/master/java/kudu-client/src/test/java/org/kududb/client/ITScannerMultiTablet.java#L120
In this failure we detected four tablets, but I checked the logs and we only talked to 3 tablets (and sent 3 scans).
So it could mean that there's a bug in sorting.
was (Author: jdcryans):
This is worrying. We have a hash-partitioned table into 3 tablets, with a total of 20k rows, and we do a sorted scan. We "detect" going across tablet boundaries by simply checking if we're seeing a row key go down: https://github.com/cloudera/kudu/blob/master/java/kudu-client/src/test/java/org/kududb/client/ITScannerMultiTablet.java#L120
In this failure we detected for tablets, but I checked the logs and we only talked to 3 tablets (and sent 3 scans).
So it could mean that there's a bug in sorting.
> org.kududb.client.ITScannerMultiTablet is flaky
> -----------------------------------------------
>
> Key: KUDU-1531
> URL: https://issues.apache.org/jira/browse/KUDU-1531
> Project: Kudu
> Issue Type: Bug
> Affects Versions: 0.9.1
> Reporter: Mike Percy
>
> Flaky test observed on Jenkins at http://104.196.14.100/job/kudu-gerrit/2439/BUILD_TYPE=ASAN/console
> Log snippet:
> {code}
> 18:09:38 T E S T S
> 18:09:38 -------------------------------------------------------
> 18:09:38 Running org.kududb.client.ITScannerMultiTablet
> 18:11:32 Tests run: 3, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 23.529 sec <<< FAILURE! - in org.kududb.client.ITScannerMultiTablet
> 18:11:32 testSortResultsByPrimaryKey(org.kududb.client.ITScannerMultiTablet) Time elapsed: 0.535 sec <<< FAILURE!
> 18:11:32 java.lang.AssertionError: expected:<3> but was:<4>
> 18:11:32 at org.junit.Assert.fail(Assert.java:88)
> 18:11:32 at org.junit.Assert.failNotEquals(Assert.java:743)
> 18:11:32 at org.junit.Assert.assertEquals(Assert.java:118)
> 18:11:32 at org.junit.Assert.assertEquals(Assert.java:555)
> 18:11:32 at org.junit.Assert.assertEquals(Assert.java:542)
> 18:11:32 at org.kududb.client.ITScannerMultiTablet.testSortResultsByPrimaryKey(ITScannerMultiTablet.java:127)
> 18:11:32
> 18:11:33 Running org.kududb.client.ITClient
> 18:12:35 Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 62.177 sec - in org.kududb.client.ITClient
> 18:12:35
> 18:12:35 Results :
> 18:12:35
> 18:12:35
> 18:12:35
> 18:12:35 Flaked tests:
> 18:12:35 org.kududb.client.ITScannerMultiTablet.testSortResultsByPrimaryKey(org.kududb.client.ITScannerMultiTablet)
> 18:12:35 Run 1: ITScannerMultiTablet.testSortResultsByPrimaryKey:127 expected:<3> but was:<4>
> 18:12:35 Run 2: PASS
> 18:12:35
> 18:12:35
> 18:12:35 Tests run: 3, Failures: 0, Errors: 0, Skipped: 0, Flakes: 1
> 18:12:35
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)