[JBoss JIRA] Commented: (JBAS-3071) MemoryLeak (redeployment) on EJB3
by Clebert Suconic (JIRA)
[ http://jira.jboss.com/jira/browse/JBAS-3071?page=comments#action_12340349 ]
Clebert Suconic commented on JBAS-3071:
---------------------------------------
No.. this is not related to the MBeanProxy...
This is a redeployment leakage. I would need to work with BillB (or someone else he assigns) on this one.
There are some containers holding references to datatypes referencing classLoaders (reflection?). This will need some refactoring/decision on how to do this.
> MemoryLeak (redeployment) on EJB3
> ---------------------------------
>
> Key: JBAS-3071
> URL: http://jira.jboss.com/jira/browse/JBAS-3071
> Project: JBoss Application Server
> Issue Type: Bug
> Security Level: Public(Everyone can see)
> Components: EJB3
> Affects Versions: JBossAS-4.0.4.CR2
> Reporter: Clebert Suconic
> Assigned To: Bill Burke
> Priority: Critical
> Fix For: JBossAS-5.0.0.Beta, JBossAS-4.0.5.CR1
>
>
> This might be caused by JBAS-3016:
> But I'm seeing this strong reference:
> org.jboss.mx.loading.UnifiedClassLoader3@33411906
> !--- org.jboss.ejb3.stateless.StatelessContainer@29610827
> We should make sure EJB3 packages are not leaking.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.jboss.com/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
18 years, 5 months
[JBoss JIRA] Reopened: (JBAS-2225) FCI.getTargets() should be Immutable
by Brian Stansberry (JIRA)
[ http://jira.jboss.com/jira/browse/JBAS-2225?page=all ]
Brian Stansberry reopened JBAS-2225:
------------------------------------
Serializing the new ImmutableArrayList in a proxy will break clients that don't have that class, so all the proxy writeObject/External methods need to serialize a clone.
Also, ImmutableArrayList needs to use delegation, not just subclassing. If it's a subclass, the internal array gets copied as part of the c'tor; if we're going to do that it makes as much sense to just return a defensive copy of the original ArrayList and not create a separate class.
> FCI.getTargets() should be Immutable
> ------------------------------------
>
> Key: JBAS-2225
> URL: http://jira.jboss.com/jira/browse/JBAS-2225
> Project: JBoss Application Server
> Issue Type: Sub-task
> Security Level: Public(Everyone can see)
> Components: Clustering
> Affects Versions: JBossAS-4.0.3RC2
> Reporter: Adrian Brock
> Assigned To: Brian Stansberry
> Fix For: JBossAS-4.0.5.CR1
>
>
> The targets returned from the FamilyClusterInfoImpl.getTargets() should not be modified,
> otherwise you can cause ConcurrentModificationException.
> To better trap people introducing errors in this area, make a subclass of ArrayList
> called ImmutableArrayList and return an implementation of this class from getTargets.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.jboss.com/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
18 years, 5 months
[JBoss JIRA] Resolved: (JBAS-3359) Implement StaleConnection checking logic in JBossJCA
by Weston Price (JIRA)
[ http://jira.jboss.com/jira/browse/JBAS-3359?page=all ]
Weston Price resolved JBAS-3359.
--------------------------------
Fix Version/s: JBossAS-5.0.0.Beta
Resolution: Done
Currently this is checked into HEAD. We need to determine with the client that requested this, how they want it applied (ie patch, version upgrade etc, etc)
> Implement StaleConnection checking logic in JBossJCA
> ----------------------------------------------------
>
> Key: JBAS-3359
> URL: http://jira.jboss.com/jira/browse/JBAS-3359
> Project: JBoss Application Server
> Issue Type: Feature Request
> Security Level: Public(Everyone can see)
> Components: JCA service
> Reporter: Weston Price
> Assigned To: Weston Price
> Priority: Minor
> Fix For: JBossAS-5.0.0.Beta
>
>
> Currently JBoss/JCA has no support for handling 'stale' connection conditions wherein a JDBC connection fails during use. The ValidConnectionChecker is only executed prior to use or in the background as of JBoss 4.0.5. Shutting of exception sorting in this situation is a bit drastic in that no exceptions would be seen as FATAL anymore. This feature would add the capablity of handling a stale connection and permit an application to retry SQL executions that may have failed under certain conditions.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.jboss.com/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
18 years, 5 months