Small ECO w/ Weblogic Timeout Exception
I was wondering if anyone’s ran into this before. We’re experiencing an increasing number of ECOs and MCOs that can’t release. They run up against our 900 second weblogic timer.
Oracle website just says to not put so many affected items but in our case we’re talking about a fraction of what we’ve seen Agile do with other ECOs and MCOs. If anyone knows what’s causing this issue or if increasing the weblogic timer will work, please let me know. I’ve seen this even going as small as an MCO with a few affected items being unable to be released.
The specific APIException is:
java.rmi.RemoteException: EJB Exception:; nested exception is:
java.rmi.RemoteException: EJB Exception: ; nested exception is:
com.agile.util.exception.CMDebugException: Transaction BEA1-2157941ABC2E325E73A5 not active anymore. tx status = Rolled back. [Reason=weblogic.transaction.internal.TimedOutException: Transaction timed out after 899 seconds
Is there any custom JAva or Groovy PX that runs during the release of the ECO/MCO? I see in the stacktrace this kind of “Mass Update – Change Objects” that seems not to be an Agile one.
In case please share the code or have a look at it in order to check if there is any issue.
Good question – it’s just a PX that attempts to release the ECO/MCO. The code is simple and has worked on other changes. Essentially it gets a change number as a scheduled event, defines next status as “released” and attempts to change status using the normal API (change.changeStatus()). There’s a try block that’s catching the API Root Cause and stack trace as I’ve posted above.
That’s why I ruled out the script itself having any issues. I’ve eliminated it down to where it’s not this script or any other script as I’ve disabled the one other one running against this event.
The “Mass Update – Change Objects” is just a temp script at the moment running to try and release the troubling ECO/MCOs as a scheduled event.
<br /><p>IChange change = session.getObject(IChange.OBJECT_TYPE, "MCO06662")</p><p>IStatus released = (IStatus) change.getWorkflow().getStates(StatusConstants.TYPE_RELEASED)</p><p>change.changeStatus(released, false, "Pushing to complete using automatic script", false, false, null, null, null, null, false)</p><br />
Is there anything else in the stack trace that seems unusual?
I believe the error message when attempting as a web client user was “action did not complete within the stipulated time” (or something like that).
We did figure out, just for anyone’s knowledge, that when Agile goes through the ECO/MCO release process, it conducts a BOM Explosion report (use Oracle SQL) to update the BOM object numbers (i.e., BOM serial number entries). We were releasing MCOs and ECOs with 20+ BOM levels. Because this took over 15 minute for us, weblogic failed.