[
https://issues.apache.org/jira/browse/ODE-817?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Terry Mueller updated ODE-817:
------------------------------
Attachment: ODE-817.patch
Attached patch
> INVOKE_CHECK jobs are not cleaned up
> ------------------------------------
>
> Key: ODE-817
> URL: https://issues.apache.org/jira/browse/ODE-817
> Project: ODE
> Issue Type: Bug
> Components: BPEL Runtime
> Affects Versions: 2.0-beta2
> Environment: Path: .
> URL: http://svn.apache.org/repos/asf/ode/trunk
> Repository Root: http://svn.apache.org/repos/asf
> Repository UUID: 13f79535-47bb-0310-9956-ffa450edef68
> Revision: 927259
> Node Kind: directory
> Schedule: normal
> Last Changed Author: vanto
> Last Changed Rev: 915457
> Last Changed Date: 2010-02-24 01:57:37 +0800 (Wed, 24 Feb 2010)
> Reporter: Terry Mueller
> Attachments: ODE-817.patch
>
>
> After running the system for a while, we now have 1000's of entries in the
> ode_job table with type = 'INVOKE_CHECK'.
> The handler of these events doesn't appear to invoke jobCompleted(). The
> following patch appears to work. Note that the null check is also required
> for longer timeouts.
> Index: engine/src/main/java/org/apache/ode/bpel/engine/BpelServerImpl.java
> ===================================================================
> --- engine/src/main/java/org/apache/ode/bpel/engine/BpelServerImpl.java
> (revision 927259)
> +++ engine/src/main/java/org/apache/ode/bpel/engine/BpelServerImpl.java
> (working copy)
> @@ -555,11 +555,12 @@
> if (__log.isDebugEnabled()) __log.debug("handleWorkEvent:
> InvokeCheck event for mexid " + j.getMexId());
>
> PartnerRoleMessageExchange mex =
> (PartnerRoleMessageExchange) getMessageExchange(j.getMexId());
> - if (mex.getStatus() == MessageExchange.Status.ASYNC ||
> mex.getStatus() == MessageExchange.Status.ACK) {
> + if (mex != null && (mex.getStatus() ==
> MessageExchange.Status.ASYNC || mex.getStatus() ==
> MessageExchange.Status.ACK)) {
> String msg = "No response received for invoke (mexId=" +
> j.getMexId() + "), forcing it into a failed state.";
> if (__log.isDebugEnabled()) __log.debug(msg);
>
> mex.replyWithFailure(MessageExchange.FailureType.COMMUNICATION_ERROR, msg,
> null);
> }
> + process._contexts.scheduler.jobCompleted(jobInfo.jobName);
> return;
> }
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
