[SOLVED] Why I need to mount ZFS volume manualy for container after "update-initramfs" usage ?

[LeDidou]

Hello

I have a problem on my Proxmox server suddenly some days ago , impossible to start containters.

It's because ZFS corresponding volume is not mount ! (I don't know when it should be mount)

The workaround I have found is to run "zfs mount -O -a" manually as root at each reboot. It works

Do you have idea of what I have done to break my config ( some config file has changed to configure PciPassThrough)

And how to restore initial Proxmox config ?

Thanks

Didier

 

[Stoiko Ivanov]

maybe it's a missing cachefile
Try the following workaround: https://forum.proxmox.com/threads/update-broke-lxc.59776/#post-277303

I hope this helps!

 

[LeDidou]

It was the problem !

I resume the solution for me

(I have a zfs pool named "mpool")

> zpool set cachefile=/etc/zfs/zpool.cache mpool
> update-initramfs -u -k all

I suppose during Pci Passthrough configuration an "update-initramfs" has broken the initial configuration...

Thanks a lot

 

[Stoiko Ivanov]

Glad that worked!

Short question - how did you create mpool initially? and did it work after a reboot at some point in the past?

 

[lilp]

That don't erase any data?

 

[Stoiko Ivanov]

the 2 commands:

zpool set cachefile=/etc/zfs/zpool.cache $POOLNAME
update-initramfs -u -k all

should not erase any data

run the zpool command for all configured pools!