[infinispan-issues] [JBoss JIRA] Commented: (ISPN-832) TimeoutException with 2 or more nodes when there are Entity inserts, updates or removes.

[Galder Zamarreño (JIRA)]

    [ https://issues.jboss.org/browse/ISPN-832?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12570364#comment-12570364 ] 

Galder Zamarreño commented on ISPN-832:
---------------------------------------

The problem is that the timestamp cache should be replicated asynchronously but it's replicated synchronously due to state transfer being enabled. So, effectively this is putting strain in the timestamp cache which is very chatty.

I'll create a JIRA in the mean time to deal with this better, but the quick solution is to disable state transfer in update timestamps and instead use a clustered cache loader. I'll try this out now.

> TimeoutException with 2 or more nodes when there are Entity inserts, updates or removes.
> ----------------------------------------------------------------------------------------
>
>                 Key: ISPN-832
>                 URL: https://issues.jboss.org/browse/ISPN-832
>             Project: Infinispan
>          Issue Type: Bug
>          Components: State transfer
>    Affects Versions: 4.2.0.CR3
>         Environment: Windows XP + Hibernate 3.6.1-SNAPSHOT + OpenEjb 3.2-SNAPSHOT
>            Reporter: Cyrille Charron
>            Assignee: Galder Zamarreño
>         Attachments: InfiniSpanTest.zip
>
>
> Hello,
> I have some difficulties to run more than one Node 
> with Infinispan as a JPA Level2 cache (default configuration)
> with Hibernate as JPA provider and 
> with OpenEjb as JEE container.
> There are a lot of Locks between Nodes when I try to insert entities,
> even if each Node insert entities with a unique ID.

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira