[JBoss jBPM] - Re: Deploying jbpm.war(3.1.3) to jboss as 4.0.5GA
by ngtdave
anonymous wrote : As I understand a ClasscastException is raised at line 435 in the class SAXReader,
| this exception is caught at line 468 and again a DocumentException is thrown at line 484:
I think you are right, that is what I am finding also. Is there any way to get the details of the stack trace/class cast exception?
I thought it might be that jbpm is packaged with a different dom4j jar than jboss 4.0.5, but looking at the manifests of server\default\lib\dom4j.jar and jbpm.war\WEB-INF\lib\dom4j-1.6.1.jar, they look to both be Implementation-Version: 1.6.1
Of course it could be a class loading issue, one was loaded from the container and the other from the war... still looking into it, but thought I would "think out loud" to see if we can get this figured out faster.
Dave
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3982074#3982074
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3982074
19 years, 5 months
[JBoss Seam] - Nested conversation question
by Ci
Hi!
I tried to use nested conversation. I took an Issue example and made my CRUD after it. I have a set of Clients, every Client has a set of Addresses. I wanted to organize a CRUD around a selected Client and a nested CRUD around a selected Address.
But there are situations where this approach fails. When a nested conversation is created (inside a parent one) and destroyed after the request is accomplished, there are an unlocked nested conversation and locked parent one, and the latter is locked by the thread of the last request, say http-0.0.0.0-8080-11.
Then everything is perfect until I do @End of the nested conversation. A different thread is on the stage, say http-0.0.0.0-8080-8. It destroys the nested conversation, but cannot unlock the parent, so it stays locked to http-0.0.0.0-8080-11 (normal case, ReentrantLock does everything correctly)
Now any request in any thread different from http-0.0.0.0-8080-11 will fail to restore the parent conversation, because Manager.java:536 cannot perform ce.lock(). So, I have a "No stored conversation, or concurrent call to the stored conversation" message, which breaks my understanding how it should work.
I have several workstation to test this behaviour. Of course, if tomcat creates few threads, say 3, I can miss the error for rather long time. But if there are 100 threads under tomcat, it fails immediately.
Seam is from cvs (today), JBoss 4.0.5.GA.
Please clarify what is wrong with my idea of nested conversations...
(By the way, Issues example works exactly the same way, falling down when I try to view a comment which is just created.)
I consider getting rid of them at all, but it seems to be very elegant approach - I'd better use it instead of any workarounds...
Thank you in advance.
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=3982065#3982065
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=3982065
19 years, 5 months