Yes, it's an odd one eh. I'll let you know if we get anything out of premier support.

My last action was to fail over the instances due to the hotfix install. PLE seemed to be more stable at first, but I'm seeing a similar pattern to what I was seeing with the instance on the original node. Since the PLE crashes have followed the instance it does not appear to be a hardware problem.

I've set up an extended events session to see if I can pin down PLE crashes to specific queries. Unfortunately the EE GUI is awful. Nice in terms of UI functionality, but it routinely locks up SSMS for... well... I don't know how long, I terminate the program after 3 or 4 minutes. Awful awful awful.

I also see batch compltion events with an indicated duration in the hundreds of thousands, which I know for a fact is wrong, because the application sending the calls has a 30 second timeout.