[infinispan-issues] [JBoss JIRA] (ISPN-3780) CLONE - InvalidateL1Command during ST should not cancel writing the entry by ST in nontx

[William Burns (JIRA)]

    [ https://issues.jboss.org/browse/ISPN-3780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12929045#comment-12929045 ] 

William Burns commented on ISPN-3780:
-------------------------------------

{quote}The topology change is started on A only after the invalidation has been sent to C.{quote}

Yeah it appears that the new topology was only installed on C before the invalidation arrived and A doesn't get the new topology until after the put command is completed.

Now normally I would think the topology check in the entryCommit in EntryWrappingInterceptor would then throw an error.  However the topologyId is always a -1 for InvalidateL1Commands.  It is never set in StateTransferInterceptor and the setting of the topology doesn't work in RpcMangerImpl because the command is wrapped in a CacheRpcCommand which doesn't implement TopologyAffectedCommand.  I think the easiest fix is to move the if check higher up in RpcManagerImpl, but I will get a test going first to confirm this.  Also I will try to fix the issue I found with the LastChanceInterceptor which this may also fix.
                
> CLONE - InvalidateL1Command during ST should not cancel writing the entry by ST in nontx
> ----------------------------------------------------------------------------------------
>
>                 Key: ISPN-3780
>                 URL: https://issues.jboss.org/browse/ISPN-3780
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Distributed Cache
>    Affects Versions: 5.3.0.Final
>            Reporter: Radim Vansa
>            Assignee: William Burns
>            Priority: Critical
>              Labels: 620
>             Fix For: 6.1.0.Final
>
>
> When a node which is about to receive the entries for some segment receives InvalidateL1Command, this puts the key into StateConsumer.updatedKeys. After the entries for ST are received, the entry which was invalidated is not updated -> this result in losing the entry.
> Btw., in EntryWrappingInterceptor.visitInvalidateL1Command the trace logs all looked up entries for each entry - this causes some performance problems when tracing is on and there are many invalidated entries. Please, do this only once.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira