[
https://issues.apache.org/jira/browse/RATIS-303?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16605535#comment-16605535
]
Shashikant Banerjee commented on RATIS-303:
-------------------------------------------
Thanks [~szetszwo], for the review. patch v1 moves
TestRaftStateMachineException tests to ratis-server and adds the subclasses for
the rpcs.
...Could you describe how the patch fix NullPointerException?
The Null pointer exception was caused because in the TestStateMachine, a fake
exception is thrown during preAppendTransaction. This leads to leader stepping
down so subsequent client call with leader set to null fails with null pointer
exception. Since the single cluster instance was shared among all the tests ,
leader being set to null intermittently lead to the failure of other tests as
well.
The exception is addressed by waiting for the leader to come up and sending the
next request to the proper leader after the stateMachine exception is thrown in
testRetryOnExceptionDuringReplication.
> TestRaftStateMachineException is failing with NullPointerException
> ------------------------------------------------------------------
>
> Key: RATIS-303
> URL: https://issues.apache.org/jira/browse/RATIS-303
> Project: Ratis
> Issue Type: Bug
> Components: server
> Affects Versions: 0.3.0
> Reporter: Mukul Kumar Singh
> Assignee: Shashikant Banerjee
> Priority: Major
> Labels: ozone
> Fix For: 0.3.0
>
> Attachments: RATIS-303.02.patch
>
>
> TestRaftStateMachineException is failing with the following exception
> {code}
> [ERROR]
> testRetryOnExceptionDuringReplication[2](org.apache.ratis.statemachine.TestRaftStateMachineException)
> Time elapsed: 0.001 s <<< ERROR!
> java.lang.NullPointerException
> at
> org.apache.ratis.statemachine.TestRaftStateMachineException.testRetryOnExceptionDuringReplication(TestRaftStateMachineException.java:139)
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
