geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jakob Færch (JIRA) <...@geronimo.apache.org>
Subject [jira] Created: (GERONIMO-1412) Create followed by findByPrimaryKey in same transaction ctx, caller in another jar: FindByPrimaryKey fails
Date Mon, 02 Jan 2006 14:35:01 GMT
Create followed by findByPrimaryKey in same transaction ctx, caller in another jar: FindByPrimaryKey
fails
----------------------------------------------------------------------------------------------------------

         Key: GERONIMO-1412
         URL: http://issues.apache.org/jira/browse/GERONIMO-1412
     Project: Geronimo
        Type: Bug
  Components: OpenEJB  
    Versions: 1.0    
    Reporter: Jakob Færch


During adventure builder deployment, I'm running into a problem that looks a lot like the
GERONIMO-598 Jira issue.

The setup is as follows (from the Adventure Builder 1.0.3 sample application):
opc.ear contains opc-ejb.jar and processmanager-ejb.jar
opc-ej contains
  WorkFlowManagerBean (Message Driven Bean)
processmanager-ejb contains
  ProcessManagerSB (SB)
  ManagerBean (CMP Entity Bean)

all methods involve have trans-attribute Required.

The WorkFlowManagerBean invokes (in the same transaction context) two methods on the ProcessManagerSB,
first create, which calls create on the   ManagerBean home, then updateStatus, which does
a findByPrimaryKey on the ManagerBean home (followed by a setStatus on the bean instance returned).

The findByPrimaryKey fails; I've verified that the same primary key is used on the create
and the findByPrimaryKey calls.

If I change the trans-attribute on the two methods on ProcessManagerSB to RequiresNew, everything
works fine, and I can see the ManagerBean entity persisted in the database.

Can anyone figure out what might be going on? The only difference from GERONIMO-598 I notice
is the call across jar files. Maybe someone with more insight into OpenEJB can tell, if this
could have any importance?

I've looked into the no-cache-flush mentioned in the Jira, but as far as I can tell, it would
only make things worse.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
   http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
   http://www.atlassian.com/software/jira


Mime
View raw message