Disk I/O Is Overloaded on Zabbix Server

[dantelemetry]

Hi

I recently upgraded my Zabbix server from 3.4 to 5.2 and in the process, upgraded Debian 9 to Debian 10. Ever since this upgrade, I get Disk I/O Is Overloaded triggers on the Zabbix server and sometimes during this, it also reports that it has lost connection with my hosts.

I'm running Debian 10 on a Hyper-V VM with 4GB of RAM. Database is MySQL Version 15.1 Distribution 10.3.27-MariaDB.

Please see the screenshots attached showing history cache, history syncer and CPU IOWAIT time. You can see we had an instance of this just after 3:30am and then again this afternoon.

We've increased the polling intervals on our templates which seemed to solve the problem for a time, but it seems to be doing it again.

I'd appreciate any help with this please.

Dan

[dantelemetry]

Just wanting to bump this as its causing issues with our Zabbix system as its causing rogue call outs to technicians.

[yurtesen]

Does it happen everyday with 12h intervals?
Did you check your VM host stats to see there is increase in disk IO? and this VM is the cause of it?
It is hard to say much by looking at few graphs. You could also try to enable/check slowquery logs of MySQL/MariaDB maybe.

[dantelemetry]

Hi Yurtesen

Thanks for your reply. Its more random than 12h intervals. There isn't any monitoring on the Hyper-V host at the moment.

I'll have a look at enabling slow query logs on MySQL.

[yurtesen]

If you are not monitoring the host, how can you know that this is not due to some other VM causing high load on host and effecting Zabbix indirectly?