[ 
https://issues.apache.org/jira/browse/DERBY-3632?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12603463#action_12603463
 ] 

V.Narayanan commented on DERBY-3632:
------------------------------------

I ran tests on this patch and noticed multiple failures.

1) I had a "java.net.SocketException: Connection reset" initially
2) A whole lot of java.security.PrivilegedActionException: 
javax.management.InstanceNotFoundException: 
org.apache.derby:type=NetworkServer,system=c013800d-011a-69a4-8c8a-ffff861961bb
3) And quite a few  Failed to start database 'singleUse/oneuse49', see the next 
exception for details.

These might as well be because I missed a runic option while starting the 
tests. I will run the tests again and shall report back with the test results.

I am careful with this patch because it primarily a timing change in starting 
the replication master. The failures might also be due to timing changes 
induced by this patch (Not sure about this one!).

Ole, Can you please confirm that the tests passed for you with the attached 
patch?



> Replication tests must ensure stable replication state has been reached 
> before attempting further connection or new replication commands.
> -----------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: DERBY-3632
>                 URL: https://issues.apache.org/jira/browse/DERBY-3632
>             Project: Derby
>          Issue Type: Bug
>          Components: Replication, Test
>    Affects Versions: 10.4.1.4, 10.5.0.0
>         Environment: All
>            Reporter: Ole Solberg
>            Assignee: Ole Solberg
>            Priority: Minor
>         Attachments: derby-3632_p1.diff.txt, derby-3632_p1.stat.txt
>
>
> When executing replication commands (startslave, startmaster, stopmaster, 
> stopslave, failover) tests must make sure that correct replication state has 
> been reached before attempting further connection to the master and slave 
> databases.
> This causes intermittent errors in replication tests.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to