[SOLVED] VM failed on "io-error"

N

nsc

Renowned Member
Jul 21, 2010
42
1
73
Hello,

We used a VM based on Ubuntu 24.04 on Proxmox 8.4, with a second disk of 25Tb XFS.
The VM was used as a Linux Veeam Repository (Veeam 12)

Saturday after 1 month without any error, the VM failed in error.

Since then, we have been unable to remount the disk, because there are unrepairable XFS errors, XFS Repair failed and the VM "hangs" in io-error

qemu-img check finds no errors.


The Proxmox host is based on a software RAID5 with an EXT4 partition, so there are no errors on this side either.
This server had been used for several years in Proxmox 6 with the same configuration (RAID5 MDAM and Windows VM in REFS).
We just reinstalled it in 8.4 and made this new VM.

Perhaps someone has an idea?

Nsc
 
OK the RAID5 has not failed but i found a lot of "mdadm mismatch sector in range" after a check ...

I don't really understand why the cloud provider who produced the Proxmox template for their server didn't include a check in the template to provide a monthly analysis.

We're going to set this up ourselves.
 
You must log in or register to reply here.
Top Bottom