[jboss-dev-forums] [Design of JBoss Portal] - Re: portal clustering problem when using optimistic locking

[galder.zamarreno@jboss.com]

Prabhat,

It's true that pessimistic locking for 2nd level cache is currently our preferred configuration, hence it's the default for EJB3 2nd level cache in EAP/AS 4.2+ but I dunno the exact details on why PL was chosen over OL. I'm gonna forward the forum link to Brian who I think might be able to give us further information on why PL was chosen over OL.

In the mean time, I don't have enough info to determine the root cause of your issues and so I'd like you to follow this instructions, repeat the test and attach a zip file with the resulting files to the JIRA.

- Add thread logging information and enable TRACE for org.jboss.cache package. See [1]

- Generate thread dumps in both nodes when you're unable to access that page.

Once you've done this, could you try with the latest EAP 4.2.x CP which is EAP 4.2.0.GA_CP03? This is our latest release for customers which has newer bug fix releases for JBoss Cache and Hibernate amongst other things.

[1] Add to conf/jboss-log4j.xml:
  |    <category name="org.jboss.cache">
  |       <priority value="TRACE" />
  |    </category>
  | 
  | Change ConversionPattern for server.log file appender to full pattern:
  | <param name="ConversionPattern" value="%d %-5r %-5p [%c] (%t:%x) %m%n"/>

View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4158663#4158663

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4158663