Comment 18 for bug 1327763

Bert Hutzler (bhutzler) wrote :

As i already stated, we run a 5 node cluster with databases on SSDs, dedicated switch for sync and doing writes mainly on only one node.
Actually we're running Percona-XtraDBCluster 5.6.22 25.8.978.el6,
and still some nodes crash randomly but rather rarely (about once or twice a week)
I handled this with a script running every minute and restarting mysql if it's crashed...

We're productive and this runs ok for us, although we'd be glad of course getting totally rid of the random crashes...
And we're planning to update this week to the actual release 5.6.24-25.11.1.el6 this week.

I'll report changes, hopefully improvements after updating...