[jira] [Commented] (SOLR-13065) Harden TestSimExecuteActionPlan

["Jason Gerlowski (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SOLR-13065?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16719267#comment-16719267 ] 

Jason Gerlowski commented on SOLR-13065:
----------------------------------------

At first glance, this looks like a similar problem to what I recently saw in SOLR-13045.  The test fails in a {{waitForState}} block, but there's some indication that we're using an outdated (cached?) copy of the clusterstatus info.

Here's a partial stack from a recent failure I got:

{code}
  [beaster]   2> NOTE: reproduce with: ant test  -Dtestcase=TestSimExecutePlanAction -Dtests.method=testIntegration -Dtests.seed=18902C9108C137F1 -Dtests.slow=true -Dtests.badapples=true -Dtests.locale=es-GT -Dtests.timezone=Asia/Rangoon -Dtests.asserts=true -Dtests.file.encoding=UTF-8
  [beaster]   2> 24745 INFO  (simCloudManagerPool-112-thread-8) [    ] o.a.s.c.CloudTestUtils -- wrong number of active replicas in slice shard1, expected=1, found=2
  [beaster] [12:26:46.105] FAILURE 2.13s | TestSimExecutePlanAction.testIntegration {seed=[18902C9108C137F1:7163CC06353074F9]} <<< 
  [beaster]    > Throwable #1: java.lang.AssertionError: Timed out waiting for replicas of collection to be 2 again
  [beaster]    > Live Nodes: [127.0.0.1:10016_solr]
  [beaster]    > Last available state: DocCollection(testIntegration//clusterstate.json/444)={
 ...
  [beaster]    >  at __randomizedtesting.SeedInfo.seed([18902C9108C137F1:7163CC06353074F9]:0)
  [beaster]    >  at org.apache.solr.cloud.CloudTestUtils.waitForState(CloudTestUtils.java:70)
  [beaster]    >  at org.apache.solr.cloud.autoscaling.sim.TestSimExecutePlanAction.testIntegration(TestSimExecutePlanAction.java:200
...
  [beaster]    > Caused by: java.util.concurrent.TimeoutException: last ClusterState: znodeVersion: 445
{code}

Note the different reported "last" clusterstate versions.  We see that there's a clusterstate.json version 445, but the failing assertion only has 444.  That's not to say definitively that version 445 would pass the assertion, but it's a place to start. 

> Harden TestSimExecuteActionPlan
> -------------------------------
>
>                 Key: SOLR-13065
>                 URL: https://issues.apache.org/jira/browse/SOLR-13065
>             Project: Solr
>          Issue Type: Test
>      Security Level: Public(Default Security Level. Issues are Public) 
>    Affects Versions: master (8.0)
>            Reporter: Jason Gerlowski
>            Assignee: Jason Gerlowski
>            Priority: Major
>
> TestSimExecuteActionPlan is a serial offender in our failed Jenkins jobs.  Would like to look into improving it.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org