This content has been marked as final. Show 3 replies
We have faced this issue before in our environment as well.
We have observed that this happens when a request is received by a server after a request that was (/got) terminated immaturely.
Which in-turn might have been caused either by the client program quitting inappropriately, or by the server received a signal, which put the process's queue in an unstable state.
This later thing can happen when the server does not handle inappropriate request termination properly, e.g might be holding on some resources which it needs to release
or is trying to forward/handle a corrupt message because the premature termination of the client process.
You may be do a echo bbc|tmadmin to clean up stale/unstable shared resources and restart the tuxedo server.
The documentation of the message (LIBTUX_CAT:669) suggest that you should focus your investigation on the queue in question (id=3003151?)
Hope this helps.
Thanks and Regards,
Yes, even I had done the same as a workaround. Bounced the tux & connectivity was fine. I just wanted to ensure that is this something related to the code or any tuxedo patch that needs to be applied.
Is this problem repeatable? In general it means that Tuxedo tried to use a System V IPC queue that no longer exists. This can happen as others say if a server or client dies unexpectedly.
Oracle Tuxedo Chief Architect