[jboss-jira] [JBoss JIRA] (AS7-4786) MSC controller hangs when loading an ear with multiple presistance units on a single threaded machine

Tue May 15 01:42:17 EDT 2012

    [ https://issues.jboss.org/browse/AS7-4786?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12692828#comment-12692828 ] 

Stuart Douglas commented on AS7-4786:
-------------------------------------

I suspect the root cause of this would be MSC thread pool starvation, caused by org.jboss.as.jpa.service.PersistenceUnitServiceImpl starting in a MSC thread, instead of doing an async start. 

Can you try starting the server with -Dorg.jboss.server.bootstrap.maxThreads=30 on the single core machine and see if this works? 

> MSC controller hangs when loading an ear with multiple presistance units on a single threaded machine
> -----------------------------------------------------------------------------------------------------
>
>                 Key: AS7-4786
>                 URL: https://issues.jboss.org/browse/AS7-4786
>             Project: Application Server 7
>          Issue Type: Bug
>         Environment: Centos 6, single core lib-virt VM
>            Reporter: Wouter De Borger
>            Assignee: David Lloyd
>              Labels: hibernate, naming, race
>
> When deploying an EAR with two persistence units on single core machines, the server hangs.
> When increasing the number of cores to two, the same server just works. 
> I traced the bug to the method in org.jboss.as.naming.WritableServiceBasedNamingStore.unbind 
> The container attempts to register "hibernate/SessionFactory" twice.
> The first time it works, the second time it hangs on the wait statement.
> I suspect a race condition in the ServiceControllerImpl.

--
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