[
https://issues.apache.org/jira/browse/HDFS-10962?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrew Wang updated HDFS-10962:
-------------------------------
Attachment: HDFS-10962.001.patch
I was unable to repro this locally after running it 100 times with load, but
looking at the code I suspect it's due to the thread pool executor being
cancelled before the badmock gets called.
So, I attempted to fix this in two ways:
* Reverse the order of badmock and goodmock so badmock gets called first
* Added a little sleep before goodmock responds to give time for the badmock to
also get called.
> TestRequestHedgingProxyProvider is flaky
> ----------------------------------------
>
> Key: HDFS-10962
> URL: https://issues.apache.org/jira/browse/HDFS-10962
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: test
> Affects Versions: 2.6.4
> Reporter: Andrew Wang
> Assignee: Andrew Wang
> Attachments: HDFS-10962.001.patch
>
>
> This test fails occasionally with an error like this:
> {noformat}
> org.apache.hadoop.hdfs.server.namenode.ha.TestRequestHedgingProxyProvider.testHedgingWhenOneFails
> Error Message
> Wanted but not invoked:
> namenodeProtocols.getStats();
> -> at
> org.apache.hadoop.hdfs.server.namenode.ha.TestRequestHedgingProxyProvider.testHedgingWhenOneFails(TestRequestHedgingProxyProvider.java:78)
> Actually, there were zero interactions with this mock.
> Stacktrace
> org.mockito.exceptions.verification.WantedButNotInvoked:
> Wanted but not invoked:
> namenodeProtocols.getStats();
> -> at
> org.apache.hadoop.hdfs.server.namenode.ha.TestRequestHedgingProxyProvider.testHedgingWhenOneFails(TestRequestHedgingProxyProvider.java:78)
> Actually, there were zero interactions with this mock.
> at
> org.apache.hadoop.hdfs.server.namenode.ha.TestRequestHedgingProxyProvider.testHedgingWhenOneFails(TestRequestHedgingProxyProvider.java:78)
> Standard Output
> 2016-09-26 15:26:02,234 WARN hdfs.DFSUtil
> (DFSUtil.java:getAddressesForNameserviceId(689)) - Namenode for
> mycluster-26780990 remains unresolved for ID nn1. Check your hdfs-site.xml
> file to ensure namenodes are configured properly.
> 2016-09-26 15:26:02,340 WARN hdfs.DFSUtil
> (DFSUtil.java:getAddressesForNameserviceId(689)) - Namenode for
> mycluster-26780990 remains unresolved for ID nn2. Check your hdfs-site.xml
> file to ensure namenodes are configured properly.
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
