[JBoss-user] [Installation, Configuration & Deployment] - Re: Startup problems

2006-04-13 Thread heismann
Just wanted to clarify sonnyboy73's last post. This exception occurs if an HTTP request is sent to the AJP connector port. This port is specified in the $SERVER_HOME/server/$CONFIG/deploy/jbossweb-tomcat55.sar/server.xml file. View the original post : http://www.jboss.com/index.html?module

[JBoss-user] [JBoss jBPM] - Re: Persisting jBPM in a stateless session bean

2006-01-10 Thread heismann
It has been pointed out to me that not closing the jbpmSession can be a bad thing in a multithreaded environment. JbpmSession's close method includes a call to popCurrentSession() which should be called or a memory leak may occur. View the original post : http://www.jboss.com/index.html?modul

[JBoss-user] [JBoss jBPM] - Re: Transaction demarcation with EJB/CMT

2006-01-10 Thread heismann
Alex, Thanks for the reply. I had forgotten about popCurrentSession() being in the close() method. So, for now I will be forced to use: jbpmSession.beginTransaction(); jbpmSession.commitTransaction(); jbpmSession.close(); Zac View the original post : http://www.jboss.com/index.html?module=b

[JBoss-user] [JBoss jBPM] - Re: Persisting jBPM in a stateless session bean

2006-01-10 Thread heismann
I've found that not closing the JbpmSession object will allow the EJB container the opportunity to close and flush the associated hibernate session. Make sense? View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3916745#3916745 Reply to the post : http://www.jb

[JBoss-user] [JBoss jBPM] - Re: Transaction demarcation with EJB/CMT

2006-01-05 Thread heismann
It appears my only mistake was in closing the jbpmSession and not allowing the system to take care of it for me. Calling the SSBs sessionContext.setRollbackOnly() had the expected result of not persisting the processInstance. Does anyone know if not closing the jbpmSession will have any adverse

[JBoss-user] [JBoss jBPM] - Re: Transaction demarcation with EJB/CMT

2006-01-04 Thread heismann
Hi Ranier, anonymous wrote : f the processInstance is being commited (persisted) that should be what you are expecting or isn't it? Yes I expect the processInstance to be committed when flush/commitTransaction is invoked. I don't think I should have to call session.flush() or jbpmSession.commi

[JBoss-user] [JBoss jBPM] - Re: Transaction demarcation with EJB/CMT

2006-01-04 Thread heismann
Rainer, I have tried explicitly flushing, this results in the changes to the process instance being persisted. I have not tried adding begin/end transaction as I'm hoping to avoid any explicit transaction demarcation code. Based upon http://hibernate.org/42.html#A7 I believe it is possible to

[JBoss-user] [JBoss jBPM] - Re: Transaction demarcation with EJB/CMT

2006-01-04 Thread heismann
Whoops, meant to hit preview, not submit. The following code is being run from a CMT SSBean. I see Hibernate running a bunch of select statements before the save method is called. After the save, however, I don't see any insert or update statements. This makes me feel that I've misconfigured

[JBoss-user] [JBoss jBPM] - Transaction demarcation with EJB/CMT

2006-01-04 Thread heismann
Hello, I've been unable to successfully configure JBPM/Hibernate to rely upon the EJB container for transaction demarcation. My Environment: JBoss 4.0.2 Hibernate 3.0.5 JBPM 3.0.2 JDK 1.5 on Solaris jbpm.properties | jbpm.log.package.org.jbpm=info | jbpm.log.package.org.jbpm.persistence

[JBoss-user] [JBoss jBPM] - Re: (jbpm+ejb session bean) need examples on how to develop

2006-01-04 Thread heismann
Hibernate Sessions and Transactions are documented here http://www.hibernate.org/42.html. Important to note this only applies to 3.1.x Hibernate versions. View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3915575#3915575 Reply to the post : http://www.jboss

[JBoss-user] [JBoss jBPM] - No TransactionManagerLookup configured

2005-12-22 Thread heismann
| [TransactionManagerLookupFactory] No TransactionManagerLookup configured (in a JTA environment, use of read-write or transactional second-level cache is not recommended) | Does anyone know what this means and how to correct it? It sounds like something is misconfigured. I'm running

[JBoss-user] [JBoss jBPM] - Re: ProcessArchiveDeployerTask

2005-12-21 Thread heismann
I was able to deploy a par using the DeployParTask. Here is a snippet from my ant script. | | | | | | | | | | | | | | | | The conte

[JBoss-user] [JBoss jBPM] - Re: ProcessArchiveDeployerTask

2005-12-20 Thread heismann
"guy.walker" wrote : I think it might be in org.jbpm.jpdl.par, the class is called ProcessArchiveDeployer | I don't think so. ProcessArchiveDeployer is not extending an Ant taskdef. I'm inclined to believe that the org.jbpm.ant.DeployParTask is what we're looking for. /** | * ant task f