Background: First, let me apologize if I get some of the details or terminology wrong, I'm still learning. I have a small unraid box that I have been using to run OPNSense as a VM and a few containers, PiHole, Uptime Kuma and maybe one or two other containers. I have no Array, I have a single NVMe configured as a ZFS cache pool where I had my containers and VM's. It has been running great for months and months, but we had a power outage while I was away and I came back to find my NVMe "unmountable: unsupported or no file system".

Unraid Version: 7.1.3

VM: OPNSense

Dockers: PiHole, Uptime Kuma

Plugins: Tailscale, ZFS Master for Unraid

Other: I passthrough 4 ethernet NICs to the OPNSense VM

Of course I have no backup. Not too sure if recovery from the NVMe is possible or if I should just rebuild everything.

Any insight would be appreciated.

smart.dump.zipbluegill-diagnostics-20250830-1133.zip

Please post output from zpool import

6 hours ago, JorgeB said:

Please post output from zpool import

Thanks for the reply, this is what I get:

"no pools available to import"

Can't be a good sign.

1 minute ago, unraid_user said:

Can't be a good sign.

It's not, post the outputs from:

fdisk -l /dev/nvme0n1

blkid /dev/nvme0n1p1

I'm prepared to rebuild it, there was no irreplaceable data, only time and effort into putting it together.
(a proper backup will be part of that new build :)

There's no partition; you can try and see if the old fs signature is still there, worth a try.

Type sfdisk /dev/nvme0n1, then type 2048 and enter and post the output; don't do anything else for now.

No ZFS signature detected. You can try the other possible sector Unraid can use; hit CTRL + C to abort, then type sfdisk /dev/nvme0n1, but now type 64 and enter and post the output.

I'm afraid there's no previous signature, suggesting the device was fully trimmed or wipefs was run on the partition before deleting that as well.

You will need to recreate the pool, if there was important data there, you could also try a file recovery app, like UFS Explorer; the free trial should show if it can recover anything.

Thanks for helping with the diagnosis, it's much appreciated. There is no irreplaceable data, just time and effort in building so no need to go through great effort/cost to do data recovery.

I do have one last question. When you say: "suggesting the device was fully trimmed or wipefs was run on the partition before deleting that as well" Could the drive have gotten into this state on it's own via hardware failure or a bad configuration or an obscure bug? If not, and the only way it could have gotten into this state is by someone typing those commands, then it implies my server was compromised. If that's the case then I need to start looking at that.

1 hour ago, unraid_user said:

Could the drive have gotten into this state on it's own via hardware failure or a bad configuration or an obscure bug?

I've seen many times devices losing a partition without an apparent reason; it typically happens after a reboot or power cycle, both with HDDs and SSDs, losing the partition and the signature is not as common, but I've also seen it before, possibly a firmware issue. This being the result of the server being compromised seems very unlikely to me.

Phew, ok thanks for your help, I really was at a loss before you chimed in.

I will rebuild this server with reliability and backup in mind.

Please consider this issue resolved. Thanks again.