Debian VM Takes down Host
-
I have been troubleshooting my 8.3 install for a few days.
After realizing that the motherboard SAS controller may be failing I am using the sata ports on the motherboard, which seems to have worked.
I ran the host without any VMs for over 24 hours with no issues. Then I started XOA and after another 24 hours everything working great. Then I started a debianVM (bookworm) and the system went down. The host shell started giving Input/Output errors on every command, and the two VMs were hung with disk errors on their xvda devices (a different ssd than the system ssd).
journalctl shows the journal stopped a couple of days ago, but /var/logs has only 8% usage so there is plenty of disk space.
I would appreciate any insight on where I might look to try to troubleshoot this. I did an fsck on the root partition and it cleared an orphaned inode, but I don't know what that means.
Dell diagnostics long test repeatedly comes back with no errors, so I think the memory is fine. Could the controller issue mean that even the SATA ports are suspect? I would hope that the diagnostics would tell me that but perhaps it is intermittent and only after running for a while.
A reboot (forced shutdown by holding down the power button) fsck of the sr (through /dev/mapper) shows no issues on the SR, but the VM consoles did show IO errors on their xvdaN disks from that SR.
What else can I do? -
@PGibson It could be the format of the drives on the SAS port and then moved to the SATA port. If the hardware is now "good" you might want to reinstall and start over with the new configuration.
-
@Andrew
Ok, thanks I will give that a try.