[ 
https://issues.apache.org/jira/browse/FLINK-26341?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17754079#comment-17754079
 ] 

Rui Fan commented on FLINK-26341:
---------------------------------

Hi [~mapohl], I have already updated ZooKeeperTestEnvironment to 
ZooKeeperExtension for ZooKeeperStateHandleStoreTest in 
https://github.com/apache/flink/pull/23196 . 

Besides this test, I still found 4 test classes:
* ZooKeeperCheckpointIDCounterITCase
* ZooKeeperCompletedCheckpointStoreITCase
* ZooKeeperJobGraphsStoreITCase
* JobManagerHAProcessFailureRecoveryITCase

I propose migrating these tests from junit4 to junit5 and removing the 
{{ZooKeeperTestEnvironment}} directly via this JIRA. Because most of developers 
aren't familiar with the ZooKeeperExtension. It's easy to forget to update the 
ZooKeeperExtension.

What do you think? If you agree to do it alone, I'd like to take it, thanks~ :)

> ZooKeeperTestEnvironment does not allow for repeated test execution
> -------------------------------------------------------------------
>
>                 Key: FLINK-26341
>                 URL: https://issues.apache.org/jira/browse/FLINK-26341
>             Project: Flink
>          Issue Type: Bug
>          Components: Runtime / Coordination
>    Affects Versions: 1.13.6, 1.14.3, 1.15.0
>            Reporter: Matthias Pohl
>            Priority: Minor
>              Labels: auto-deprioritized-major, stale-minor, test-stability
>
> {{ZooKeeperTestEnvironment.deleteAll}} fails when rerunning a test due to the 
> client not being started:
> {code}
> java.lang.IllegalStateException: Client is not started
>       at 
> org.apache.flink.shaded.curator5.org.apache.curator.shaded.com.google.common.base.Preconditions.checkState(Preconditions.java:507)
>       at 
> org.apache.flink.shaded.curator5.org.apache.curator.CuratorZookeeperClient.getZooKeeper(CuratorZookeeperClient.java:139)
>       at 
> org.apache.flink.runtime.zookeeper.ZooKeeperTestEnvironment.deleteAll(ZooKeeperTestEnvironment.java:145)
>       at 
> org.apache.flink.runtime.checkpoint.ZooKeeperCheckpointIDCounterITCase.cleanUp(ZooKeeperCheckpointIDCounterITCase.java:48)
>       at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
>       [...]
> {code}
> This behavior was observed when running various tests with 
> repeat-until-failure enabled in Intellij:
> * {{ZooKeeperCheckpointIDCounterITCase.testShutdownRemovesState}}
> * {{ZooKeeperStateHandleStoreTest.testAddAndLock}}
> * 
> {{ZooKeeperStateHandleStoreTest.testFailingAddWithPossiblyInconsistentState}}
> I'd assume that this applies in any other case as well.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to