Message flow stops for all users on a BlackBerry Enterprise Server for GroupWise due to initialization issues for one or more users

• BlackBerry® Enterprise Server 5.0 SP1 to 5.0 SP1 MR5 for Novell® GroupWise®
• DT 2412883
• DT 2412856
  

This issue is evident when the Messaging Agent has been started, or has been restarted due to hung threads, and does not initialize the SOAP Connector upon restart (no logging is present in the GWSC log file since the agent start time). Under normal circumstances, we would see something like this:

[MAGT]

[30000] (10/03 00:07:59.112):{0x1D80} [DIAG] EVENT=Register_thread, THREADID=0x1D80, THREADNAME="SCS"
[30047] (10/03 00:07:59.112):{0x1D80} Calendar synchronization has been enabled
[30049] (10/03 00:07:59.112):{0x1D80} Starting handhelds
[40000] (10/03 00:08:48.865):{0x1D80} SetupDeliveryStatus - Completed successfully
[40000] (10/03 00:08:49.053):{0x1D80} DetermineMaxCalendarID - Initialized max calendar ID for DB polling
[40000] (10/03 00:08:49.569):{0x1D80} DetermineMaxMailID - Initialized max mail ID for DB polling
[40000] (10/03 00:08:49.569):{0x1D80} GWDBNotifications::ForceTimeout -- updating last timeout
[40000] (10/03 00:08:49.569):{0x1D80} GWDBNotifications::ForceTimeout -- updating last timeout
[40000] (10/03 00:08:49.569):{0x1D80} Successfully initialized database polling thread.
[30000] (10/03 00:08:50.819):{0x1D80} [GWSC] Starting
[30000] (10/03 00:09:07.383):{0x1D80} [GWSC] Started
[30362] (10/03 00:09:07.383):{0x1D80} This agent will handle xx user(s)
[30050] (10/03 00:09:07.414):{0x1D80} All handhelds started
[50079] (10/03 00:09:07.414):{0x1D80} x user(s) failed to initialize

However in cases where one or more starter threads are hung, we see this:

[MAGT]

[30000] (11/17 09:10:32.698):{0x1628} [DIAG] EVENT=Register_thread, THREADID=0x1628, THREADNAME="SCS"
[30047] (11/17 09:10:32.698):{0x1628} Calendar synchronization has been enabled
[30049] (11/17 09:10:32.698):{0x1628} Starting handhelds
...
[20372] (11/17 09:40:32.365):{0x1628} All handhelds are not started within 30 minutes
...
[20372] (11/17 10:10:32.443):{0x1628} All handhelds are not started within 60 minutes

We do not see the GWSC being started as is seen in a normal startup. Amidst that, we see starter threads hanging:

[MAGT]

[30000] (11/17 09:10:32.698):{0x5F4} [DIAG] EVENT=Register_thread, THREADID=0x5F4, THREADNAME="StarterThread"
[30041] (11/17 09:10:32.698):{0x5F4} Starting handheld for UserDisplayName
...
[20690] (11/17 09:30:32.732):{0x16AC} Starter Thread: *** No Response *** Thread Id=0x5F4, Handle=0x95c, WaitCount=1, Starter thread - starting handheld
...
[20690] (11/17 10:20:32.778):{0x16AC} Starter Thread: *** No Response *** Thread Id=0x5F4, Handle=0x95c, WaitCount=6, Starter thread - starting handheld

With these series of events, the GroupWise Soap Connector is not initialized correctly and BlackBerry mail flow is stopped.

• Workaround 1: Correct the issue with the users' GroupWise mailbox that the agent is reporting hung threads for, and restart the controller and dispatcher services.
  
• Workaround 2: Remove the users who are experiencing hung threads from the BlackBerry Enterprise Server, and restart the controller and dispatcher services.