[
https://issues.apache.org/jira/browse/ATLAS-944?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15346136#comment-15346136
]
Nigel Jones commented on ATLAS-944:
-----------------------------------
Having changed the timeout I hit another error
Failed tests:
HiveHookIT.testAlterDBOwner:1438->assertDatabaseIsRegistered:1630->assertEntityIsRegistered:1636->waitFor:1729
Assertions failed. Failing after waiting for timeout 80000 msecs
HiveHookIT.testDropAndRecreateCTASOutput:369->assertTableIsNotRegistered:1598->assertEntityIsNotRegistered:1651->waitFor:1729
Assertions failed. Failing after waiting for timeout 80000 msecs
Tests run: 45, Failures: 2, Errors: 0, Skipped: 0
We do seem to have quite a few timeouts in our tests. I wonder if we could have
a "scaling factor". I suspect the machine I'm currently using is suffering from
slow I/O (networked drive, though should be quick via 1 Gbps network). I'll
switch to local & repeat. Not sure what fix to propose just yet.
> Build failure: GraphBackedMetadataRepositoryTest.testFullTextSearch:528
> expected:
> ---------------------------------------------------------------------------------
>
> Key: ATLAS-944
> URL: https://issues.apache.org/jira/browse/ATLAS-944
> Project: Atlas
> Issue Type: Bug
> Affects Versions: trunk
> Environment: CentOS 7.2.1511
> OpenJDK 1.7.0_101
> Maven 3.3.9
> *fresh git clone*
> Reporter: Nigel Jones
>
> Build of Atlas fails
> [INFO] Apache Atlas Repository ............................ FAILURE [13:40
> min]
> Cause:
> Results :
> Failed tests:
> GraphBackedMetadataRepositoryTest.testFullTextSearch:528 expected:<1> but
> was:<0>
> Root error:
> Running org.apache.atlas.repository.graph.GraphBackedMetadataRepositoryTest
> Tests run: 26, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 31.471 sec
> <<< FAILURE! - in
> org.apache.atlas.repository.graph.GraphBackedMetadataRepositoryTest
> testFullTextSearch(org.apache.atlas.repository.graph.GraphBackedMetadataRepositoryTest)
> Time elapsed: 1.237 sec <<< FAILURE!
> java.lang.AssertionError: expected:<1> but was:<0>
> at org.testng.Assert.fail(Assert.java:89)
> at org.testng.Assert.failNotEquals(Assert.java:489)
> at org.testng.Assert.assertEquals(Assert.java:118)
> at org.testng.Assert.assertEquals(Assert.java:365)
> at org.testng.Assert.assertEquals(Assert.java:375)
> at
> org.apache.atlas.repository.graph.GraphBackedMetadataRepositoryTest.testFullTextSearch(GraphBackedMetadataRepositoryTest.java:528)
> Looking at the code in
> ./repository/src/test/java/org/apache/atlas/repository/graph/GraphBackedMetadataRepositoryTest.java
> @Test(dependsOnMethods = "testSubmitEntity")
> public void testFullTextSearch() throws Exception {
> //todo fix this
> //Weird: with lucene, the test passes without sleep
> //but with elasticsearch, doesn't work without sleep. why??
> long sleepInterval = 1000;
> The 1000ms looks arbitary... I'm building on a 4 CPU vm which should be ok,
> but to a shared drive, perhaps I/O limited. increasing the timeout
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
