[
https://jira.jboss.org/jira/browse/JBPM-1953?page=com.atlassian.jira.plug...
]
Tom Baeyens commented on JBPM-1953:
-----------------------------------
"
Issue 1:
A lot of StaleObjectStateException are thrown by Hibernate, if a jBPM process on SOA
Platform run in a cluster.
"
==> this is expected behaviour in a cluster. the user is in that case supposed to
retry the operation.
"
Issue 2:
A StaleObjectStateException is thrown by Hibernate so easily, if the number of
JobExceutor's thread is two or more. If you set the number of thread to
./samples/quickstarts/bpm_orchestration2,
you could see the issue easily.
"
==> same thing.
"
Plan 1:
Set the name to each JobExecutor's name field in jbpm.cfg.xml, if the system using SOA
Platform is in multi-instances or in a cluster. By this plan, the name of each Job become
unique. So, the rate of this fail by optimisticLoking would become small number. but
it's not complete solution. There is still a possibility that the exception occurs.
"
==> that is indeed a requirement to get job executors to run ok on multiple nodes. but
it will not eliminate concurrent optimistic locking exceptions.
"
Plan 2:
Use JMS instead of DBMS for the store of Jobs.
"
==> that will create more problems then it solves. in fact, it will solve none.
"
Plan 3:
Select Pessimistic lock. If the system can choose the lock strategy of JobExecutor like
PessimisticLocking or OptimisticLoking, the system could choose the strategy in their
environment. But it's a kind of enhancement request to jBPM/SOA Platform.
"
==> pessimistic locking will not help either to get rid of exceptions. in a cluster,
and with competing threads trying to perform transactional operations on single process
instance, you'll always have transactions that have to roll back. even with
pessimistic locking.
Select the Lock Strategy in jbpm.cfg.xml
----------------------------------------
Key: JBPM-1953
URL:
https://jira.jboss.org/jira/browse/JBPM-1953
Project: JBoss jBPM
Issue Type: Task
Security Level: Public(Everyone can see)
Components: Core Engine
Reporter: Thomas Diesler
Assignee: Alejandro Guizar
Priority: Critical
Fix For: jBPM 3.3.2 GA
A StaleObjectStateException is thrown by Hibernate so easily, if the number of
JobExceutor's thread is two or more. If you set the number of thread to
./samples/quickstarts/bpm_orchestration2,
you could see the issue easily.
For a complete description, see the (private) document attached to JBPM-1952
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
http://www.atlassian.com/software/jira