[SOLVED] Problem with CT's not starting after upgrade to 6.4.

[McKajVah]

I'm getting a error message on all my containers after an upgrade to 6.4-5 from 6.3.x something.

"TASK ERROR: mkdir /var/lib/lxc: Structure needs cleaning at /usr/share/perl5/PVE/LXC.pm line 734."

None of my CT's will start because of this.

Any suggestions on what to do?

 

[Stefan_R]

That sounds like a file system corruption. Was the server unexpectedly powered off perhaps? This might also potentially be a disk error.

Anyhow, running fsck on your root disk (and maybe your container images) might be enough to fix the issue. What filesystem are you running as root (lsblk and mount would help)?

 

[McKajVah]

Hi.

No, the server was shut down normaly after the update and started having issues when it rebooted afterwards.

FSCK did not help and everything looks to be mounted correctly.

Had a look in the logs and looks like there is Metadata corruption detected. I probably need to try and run XFS_repair.

 

[McKajVah]

Turned out it was a SATA cable failure. Strange thing is that I didn't touch or move the machine the last year...

Had a lot of ATA errors when I hooked a monitor to it and had a look. Had to run a xfs_repair -L (forced), but luckily everything seems OK now.

Stange, very strange for a cable to just suddenly go bad inside the machine.