Problem with virtual machines after server overheating or just a coincidence?

Question

Asterisk

Hello!
It so happened that as a result of the shutdown of the cooling systems in the server room, the equipment overheated (the cisco was demonstratively buggy, driving a bunch of traffic into the port so that the provider's equipment was hung up).
Immediately after this situation, strange things began to happen with virtual machines in the north of the Dell R620, namely, virtual machines with asterisk began to fail.
Before the accident, about 500 subscriber devices "lived" on two virtual machines (2 cores, 4 GB RAM), now it is buggy on 4 dual-core and one 4-core, namely, with a slight increase in load, plugs with voice traffic begin, the error "Exceptionally long voice queue length queuing to Local/[email protected];1" and croaking begins,
The total load on the host is negligible - CPU-30%, RAM-60%, there is no load on the disks.
No crime is observed in the host logs either.
At about the same time, the space on the MariaDB virtual machine ran out, I had to emergency stop, clean it, start it over again.
In this regard, the question is what could affect the overheating of the server (according to statistics, CPU heating up to 40 degrees during the day, in peaks up to 80), t .to. Apparently there is a problem with the timings?
The iDRAC logs started three weeks after the
CTL1 crash: Controller event log: Patrol Read found an uncorrectable media error on Disk 0 in Backplane 1 of Integrated RAID Controller 1.
2016-06-18T03:07:20-0500
Log Sequence Number: 894
Detailed Description:
This event is retrieved from the controller when iDRAC storage monitoring was not running. Such events which are generated in the past are logged as informational severity.
Recommended Action:
No response action is required.

Answer 1

У вас две проблемы - первая: сервера после перегрева надо перезагрузить, и, возможно, сбросить им ошибки в BIOS.
Вторая - у вас диск 0 в бекплейне 1 начал помирать - меняйте.