Proxmox 8.2.2 Lost ZFS Pool Lost After Shutdown

[reloadtech]

Hello to all,

After an unexpectedly shutdown Proxmox server lost the ZFS Pool.
I can see from Datastore's Storage section my old ZFS Storage, but the pool is missing from the server.
Disks Health is good (I checked through Raid Controller).
All my VMs were stored in there.

Is it possible to mount my old ZFS Pool again? Is it still reachable?

Thank you all for your time.

 

[fireon]

There are ZFS commands for this. The command zpool import <pooname> is used to import a pool. See also zpool help import

 

[esi_y]

If this was one of those inexplicably corrupt situations, you may perhaps take inspiration here:
https://forum.proxmox.com/threads/z...o-error-on-import-attempt.156075/#post-712560

(Only follow the commands to the point until you HAVE TO, do not throw them all around for no reason.)

 

[reloadtech]

Thank you all for your replies,

zpool import didn't work, but zpool import -F
worked perfectly. (it's says that -F stands for recovery).

After this command my zfs pool appeared again.

My problem is that I have all my VMs on this pool. Is this a common thing? I mean to loose the pool?.

There was an sudden shutdown when it occurs to me but I read threads that others say that is happening after restart.

Is there any configuration I can do to prevent this?

Thank you for your time

 

[esi_y]

Thank you all for your replies,

zpool import didn't work, but zpool import -F
worked perfectly. (it's says that -F stands for recovery).

That's not correct, have a look at docs [1].

-f ignores that pool has not been previously exported (as daunting as it sounds it simply means that it has a marking that it was used by "another" system, which may mean you are trying to import a pool that is already active elsewhere or you might have just lost a cached value that helps your system recognise it was yours all along);

-F however means some transactions were discarded, i.e. you lost some data. If you had to use this option it dropped some last writes to the pool.

After this command my zfs pool appeared again.

My problem is that I have all my VMs on this pool. Is this a common thing? I mean to loose the pool?.

Common enough for myself not to use the most recent (potentially not well tested) versions of ZFS. But I wonder if I will be asked (by staff) now to refrain from these discussions here like it started happening on this forum in relation to raids/volumes/filesystems [2].

There was an sudden shutdown when it occurs to me but I read threads that others say that is happening after restart.

Is there any configuration I can do to prevent this?

Use a stable filesystem. Do not believe non-sense about one particular that are perpetuated on forums.

Thank you for your time

You are welcome. Whatever filesystem you use, have backups (they will [3] blame you here for not having one instantly anyways).

EDIT: If you are genuinely interested in anything ZFS, consider dedicated forum [4], it's very lacking here.

[1] https://openzfs.github.io/openzfs-docs/man/master/8/zpool-import.8.html
[2] https://forum.proxmox.com/threads/mdraid-o_direct.156036/#post-713636
[3] https://forum.proxmox.com/threads/lost-all-data-on-zfs-raid10.154843/
[4] https://zfsonlinux.topicbox.com/latest

 

[reloadtech]

Thank you,

I've been working with ESXi for many years, and over the past year, after everything that happened with VMware, I decided to try Proxmox.

I tested Proxmox for two weeks before putting it into official operation.

The reason I chose ZFS is that, compared to ext4, the VM performance speed was noticeably faster.

Regarding backups, I’ve done extensive testing with the latest version of Veeam that supports Proxmox, and I’m very satisfied. Even if I were to lose the ZFS storage, I have a second Proxmox installed to the same network, so I can restore the VMs anytime, as I send and store backups across multiple repositories.

While i was restoring backups on the second Proxmox, I discovered the zpool import command and decided to try it. I was able to bring the pool back and run the VMs again from my primary Proxmox server.

If you have any other recommendations, I’d be glad to hear them, as I'm new to the Proxmox environment.

Thank you, Best Regards
Pol Mexis

 

[esi_y]

The reason I chose ZFS is that, compared to ext4, the VM performance speed was noticeably faster.

This is a strange result, no copy-on-write filesystem with checksumming could ever perform better than RAW, even when on ext4. I can imagine this was QCOW2 but still, strange to read.

Regarding backups, I’ve done extensive testing with the latest version of Veeam that supports Proxmox, and I’m very satisfied.

Just beware of this: https://forum.proxmox.com/threads/veeam-silent-data-corruption.155212/

If you have any other recommendations, I’d be glad to hear them, as I'm new to the Proxmox environment.

Nothing that would not turn out too contentious for this forum.