Weird Mysql problems

[Erwin123]

Let me start by saying that my problem is probably not PVE related but I'm at the end of my options so any help is appreciated.

Since two months Mysql on all containers on different nodes randomly shut down.
They don'crash but just clean shut down:

100403 18:00:06 [Note] /usr/libexec/mysqld: Normal shutdown
100403 18:00:06 InnoDB: Starting shutdown...
100403 18:00:09 InnoDB: Shutdown completed; log sequence number 0 1936165485
100403 18:00:09 [Note] /usr/libexec/mysqld: Shutdown complete
100403 18:00:09 mysqld ended

I let Mysql log every query on different servers but this gave no clues.
The only thing all containers have in common is that the are on our proxmox cluster.
Our physical servers don't have this problem.

The containers are Centos 4 and Cenots 5 different Mysql versions (4 and 5 and different .xx versions) different php versions, different plesk version etc.

beancounters:

Version: 2.5
uid resource held maxheld barrier limit failcnt
134: kmemsize 16587369 40099461 9223372036854775807 9223372036854775807 0
lockedpages 0 0 262144 262144 0
privvmpages 253354 485144 524288 536788 0
shmpages 5521 9425 9223372036854775807 9223372036854775807 0
dummy 0 0 0 0 0
numproc 105 253 1024 1024 0
physpages 123316 170452 0 9223372036854775807 0
vmguarpages 0 0 524288 9223372036854775807 0
oomguarpages 124785 171831 524288 9223372036854775807 0
numtcpsock 24 192 9223372036854775807 9223372036854775807 0
numflock 9 20 9223372036854775807 9223372036854775807 0
numpty 2 3 255 255 0
numsiginfo 0 22 1024 1024 0
tcpsndbuf 497408 10543104 9223372036854775807 9223372036854775807 0
tcprcvbuf 393216 6239488 9223372036854775807 9223372036854775807 0
othersockbuf 81408 1702144 9223372036854775807 9223372036854775807 0
dgramrcvbuf 0 304384 9223372036854775807 9223372036854775807 0
numothersock 92 232 9223372036854775807 9223372036854775807 0
dcachesize 1441674 1730736 9223372036854775807 9223372036854775807 0
numfile 7418 12896 9223372036854775807 9223372036854775807 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
dummy 0 0 0 0 0
numiptent 15 15 9223372036854775807 9223372036854775807 0

The problem started after our previous firewall crashed (not sure it is in any way related), but we have port 3306 blocked for outside connections again and it still happens.

We now have a cronscript that checks mysql but thats no real solution.

Did anyone ever experience this?
Any ideas/suggestions where to search?

Thanks!