Comment 28 for bug 934377

I've started to get this issue recently as well. I've been running percona for a long time, dozens of instances and have never seen this issue before. I can't say for sure this is the only thing that has changed, but recently some portion of our application has started using temporary tables heavily.

Perhaps 2-5 temporary tables per connection, connection held for 5-30 seconds and 10+ of these connections per second.

We experienced this crash this morning, and another one 25 days ago.