We’re currently running Cloverleaf 5.8 rev2 on AIX 6.1. Yesterday we were following up on some routine maintenance and we noticed a message in the recovery database that was in state 14 and dated 11/22/2013. Both the metadata in the MH and the data in the actual message MSH are in agreement. This message has been out there in recovery since November. We had work done on the 12th of January that involved stopping the entire site and rebooting the box and the message persisted through that work. When things restarted there was no panic or anything. Messages are flowing right past the state14 message like it doesn’t even exist.
Has anyone ever seen something like this before or have an idea how a message could effectively get stranded in state 14 like that? How is it that the thread is working normally and not blocked by the state 14 message (are we now allowed multiple state 14s)? Any feel for whether or not it is safe to simply delete the message or is this a sign that we should be arranging a time to stop the entire site and reinitialize the databases?