We know it's something in the last update and are working on finding it before adding any new updates. In the changes we made to our restarter recently, we changed so that the server stops if it crashes. It will not startup again, which will from now on let us save all the data to the database, generate a core-dump if we want to that we can investigate later to track the crash. Unfortunately that wasn't possible today because the server wasn't started in the debugger (as you might have noticed by server save it didn't start at 07:00 CEST as it should have). We do expect it to crash at least one more time before we can have it addressed, but we are pretty sure we will be able to save everything so nothing gets lost in the next crash. Very sorry for the crash with data loss today, and we really hope it was the last one ever.