[infinispan-issues] [JBoss JIRA] (ISPN-1995) Uneven request balancing after node restore

[RH Bugzilla Integration (JIRA)]

    [ https://issues.jboss.org/browse/ISPN-1995?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12693682#comment-12693682 ] 

RH Bugzilla Integration commented on ISPN-1995:
-----------------------------------------------

JBoss JIRA Server <jira-update at redhat.com> made a comment on [bug 809631|https://bugzilla.redhat.com/show_bug.cgi?id=809631]

Galder Zamarreño <galder.zamarreno at redhat.com> made a comment on jira ISPN-1995

@Michal, re: node 16 test: Is the throughput stats collection right? According to them, the node that doesn't seem to recover is node0002 rather than node0003. It might be the case that the stats are only collected until view5 has been installed (http://dev39.mw.lab.eng.bos.redhat.com/~mlinhard/hyperion/run105-resi-dist-16-ER8.1-nettyfix/logs/analysis/views.html) but not view6? I can't map relative times in stats to absolute times in logs. Otherwise, I don't see any errors that would explain this.

re: node 32 tests: The stats issue looks similar, but probably it's just a false alarm. It would appear as if the stats would finish in view 12, and view 13 would not appear.

Assuming the stats are right, seems like the issue is the same: last view (view6 in 16, view 13 in 32) is not installed. This could either be because:

a) the server does not send it
b) the client does not install it

To cover all bases, repeat the test with:

1. In server, enable TRACE on {code}org.infinispan.server.hotrod.Encoders$Encoder11${code}
2. In client, enable TRACE on {code}org.infinispan.client.hotrod.impl.protocol.Codec11{code}
                
> Uneven request balancing after node restore
> -------------------------------------------
>
>                 Key: ISPN-1995
>                 URL: https://issues.jboss.org/browse/ISPN-1995
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Cache Server
>    Affects Versions: 5.1.4.CR1
>            Reporter: Tristan Tarrant
>            Assignee: Galder Zamarreño
>             Fix For: 5.1.5.CR1, 5.2.0.ALPHA1, 5.2.0.FINAL
>
>
> After a node crashes and rejoins the cluster, it does not receive client load at the same level as the other nodes.
> This issue does not affect data integrity and distribution in the cluster.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.jboss.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira