[
https://issues.apache.org/jira/browse/HIVE-11660?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Siddharth Seth updated HIVE-11660:
----------------------------------
Attachment: HIVE-11660.1.txt
Attaching patch to fix the tests. Have run 100 iterations of both on a Linux
box - where the failures are normally seen - with all of them passing.
There's some real bugs which were causing TestLlapTaskSchedulerService to fail.
The last allocateTaskRequest for a dag could've ended up being ignored.
Also in TaskScheduler, the waitQueue can be improved - filed a separate jira
for this.
[~sershe] - please review.
> LLAP: TestTaskExecutorService is flaky
> --------------------------------------
>
> Key: HIVE-11660
> URL: https://issues.apache.org/jira/browse/HIVE-11660
> Project: Hive
> Issue Type: Bug
> Reporter: Sergey Shelukhin
> Assignee: Siddharth Seth
> Attachments: HIVE-11660.1.txt
>
>
> {noformat}
> java.lang.Exception: test timed out after 10000 milliseconds
> at sun.misc.Unsafe.park(Native Method)
> at java.util.concurrent.locks.LockSupport.park(LockSupport.java:186)
> at
> java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.await(AbstractQueuedSynchronizer.java:2043)
> at
> org.apache.hadoop.hive.llap.daemon.impl.TestTaskExecutorService$TaskExecutorServiceForTest$InternalCompletionListenerForTest.awaitCompletion(TestTaskExecutorService.java:244)
> at
> org.apache.hadoop.hive.llap.daemon.impl.TestTaskExecutorService$TaskExecutorServiceForTest$InternalCompletionListenerForTest.access$000(TestTaskExecutorService.java:208)
> at
> org.apache.hadoop.hive.llap.daemon.impl.TestTaskExecutorService.testWaitQueuePreemption(TestTaskExecutorService.java:168)
> {noformat}
> Cannot repro locally. See HIVE-11642
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
