Every once in a while C/S handler threads get stuck on the GroupWise POA. It doesn't affect services on the server/client side, but when I need to restart the POA service, it hangs on trying to kill those threads. I end up having to terminate the gwpoa.exe process, which is not preferable.

It seems to be caused by a user who is using the caching mailbox function over a VPN connection. Maybe the connection doesn't terminate properly? When I look at one of the stuck threads on the POA web interface, I see the following (I left out the userID and IP address):

Access Mode Caching
Current Transaction Cursor destroy
Time Elapsed (Secs) 18184

When the thread is stuck, it's always the "Cursor destroy" transaction.

Is there anything I can do to stop this from happening, short of blocking this user from using the caching mailbox functionality? Also when it does occur, is there a way to forcibly kill the threads after a certain amount of time, when I restart the POA service?

I'm running GW 8.02 HP2 (8.0.2-92377) on Windows Server 2008 R2

Thank you.