couldn't find local-lvm storage in webui after fsck

[peyoot]

Hi, guys!

I made a stupid mistake. I don't know why my pve shutdown (maybe powerloss or something) and when I manually boot it, there's a lot of "EXT4-fs error: .........,iget: checksum invalid" 。 I didn't check if webui is availabe or not, but directly perform "fsck -c -f -y /dev/mappper/pve-root" .

It makes me lost local-lvm in webUI after reboot. I still can see the raw vm disk with "lvscan", how can I use this to bring my vm (/dev/pve/vm-100-disk-0")back?

 

[Chris]

Hi, try a pvesm scan lvm. Can you still see the VMs in the WebUI? What is the status of systemctl status pve-cluster and whats the content of cat /etc/pve/storage.cfg

 

[peyoot]

doesn't seems it have something useful,

 

[peyoot]

I finally got my vm back by manually add back local-lvm and create a new vm to use orignal vm disk. but I'd like to know the cause of checksum invalid. I look up the syslog and got where it powered off, seems after daily upgrade:


is this the cause? Can I turn off "daily PVE download activities". How to make pve update manually rather than automatically? where to do it gracefully?

 

[Chris]

I finally got my vm back by manually add back local-lvm and create a new vm to use orignal vm disk. but I'd like to know the cause of checksum invalid. I look up the syslog and got where it powered off, seems after daily upgrade:
View attachment 47288

is this the cause? Can I turn off "daily PVE download activities". How to make pve update manually rather than automatically? where to do it gracefully?

No, I would not expect the pve-daily-update.service to have anything to do with your filesystem checksum error. This service just updates the package repositories, it does not upgrade any packages. The upgrade has to be performed manually.

Filesystem corruptions occur more frequently when the system is not shut down gracefully or the disks are failing. You can check your disks SMART status.