[
https://issues.apache.org/jira/browse/LUCENE-5461?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13907517#comment-13907517
]
Hans Lund commented on LUCENE-5461:
-----------------------------------
Yes exactly:
java.lang.AssertionError: waited too long for generation 10001
at org.junit.Assert.fail(Assert.java:88)
at org.junit.Assert.assertTrue(Assert.java:41)
at
org.apache.lucene.search.TestLucene5461.testCRTReopen(TestLucene5461.java:94)
I've noticed when extracting the test that changing directory implementation
influenced how often the test failed: - using a RAMDirectory failed to trigger
the test. I've not tried with FSDirectory.
> ControlledRealTimeReopenThread waitForGeneration might sleep for
> targetMaxStaleSec instead of targetMinStaleSec
> ---------------------------------------------------------------------------------------------------------------
>
> Key: LUCENE-5461
> URL: https://issues.apache.org/jira/browse/LUCENE-5461
> Project: Lucene - Core
> Issue Type: Bug
> Components: core/search
> Affects Versions: 4.6.1
> Reporter: Hans Lund
> Priority: Minor
> Attachments: LUCENE-5461.patch, TestLucene5461.java
>
>
> If setting the tagetMinStaleSec to 0, sometimes a call to waitForGeneration
> will block for targetMaxStaleSec instead of immediately call
> maybeRefreshBlocking on the manager.
> In effect to targetMaxStaleSec cannot be set higher than acceptable blocking
> resolving of a specified generation.
>
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
