Proxmox is incredibly stable virtualization software.
If you’re having Disk IO issues, it’s typically related to hardware.
But how to troubleshoot? In this article we explain one scenario:
root@hvX:~# tail -f /var/log/syslog Apr 3 18:34:01 hvX kernel: [4956766.821761] Buffer I/O error on dev dm-10, logical block 8489400, lost async page write
From the error above, you can see there is trouble.
But what on earth is dm-10? Can you just use fdisk? In our system, we couldn’t so we sent the request to AI:
This produced this output:
root@hv09:~# ls -l /dev/mapper/ total 0 lrwxrwxrwx 1 root root 7 Feb 6 09:41 centos-root -> ../dm-8 lrwxrwxrwx 1 root root 7 Feb 6 09:41 centos-swap -> ../dm-7 crw------- 1 root root 10, 236 Feb 6 09:41 control lrwxrwxrwx 1 root root 7 Feb 6 09:41 pve-data -> ../dm-5 lrwxrwxrwx 1 root root 7 Feb 6 09:41 pve-data_tdata -> ../dm-3 lrwxrwxrwx 1 root root 7 Feb 6 09:41 pve-data_tmeta -> ../dm-2 lrwxrwxrwx 1 root root 7 Feb 6 09:41 pve-data-tpool -> ../dm-4 lrwxrwxrwx 1 root root 7 Feb 6 09:41 pve-root -> ../dm-1 lrwxrwxrwx 1 root root 7 Feb 6 09:41 pve-swap -> ../dm-0 lrwxrwxrwx 1 root root 8 Apr 3 18:18 pve-vm--107--disk--0 -> ../dm-10 lrwxrwxrwx 1 root root 7 Apr 3 18:10 pve-vm--109--disk--0 -> ../dm-6 lrwxrwxrwx 1 root root 8 Feb 6 11:07 pve-vm--113--cloudinit -> ../dm-11 lrwxrwxrwx 1 root root 7 Feb 6 09:41 pve-vm--9300--cloudinit -> ../dm-9
From here we were able to troubleshoot VM 107 to see what’s up.
To be continued.