Precursor
My cluster consisting of 2 PVE hosts and a qdevice have been happily running for months, no issues. Last weekend I saw there were updates (it's my homelab, so pve-no-subscription updates), I ran them on both hosts, rebooted and everything seemed fine. Then I tried to use an USB ethernet adapter again purely for the management interface of host 1, that worked for a few hours, then the USB adapter lost it's connection (this issue occured earlier this year, but has been running fine for months before that). I undid the bridge configuration, rebooted host 1 and...
Issue
After the reboot host 1 disconnected an external USB NVME disk (zpool ext-1tb). I unplugged it, plugged it back in, disk LED became blue (good) then after a few seconds red (not good) then no light. Safe to say, VM's who had disk images on that zpool crapped out. In Disks - ZFS the status was suspended.
What I tried
I connected the USB disk to host 2, imported the zpool and voila, the disk appeared and was showing the raw disk images of a couple of VM's of host 1, proving there's nothing wrong with the disk or USB cable. However, I could not use qm disk import , it said the source file can't be found. zfs mount told me that pool ext-1tb is mounted at /ext-1tb. When I do ls -la in that folder, there's no folder or file showing, so makes sense that qm disk import didn't find the source. With zfs send | zfs recv I was able to copy/clone a disk image to local-zfs, which resides on the internal NVME disk of host 2. In the GUI, the disk image is indeed listed now, with it's size. But qm disk import still did not find any disk image. zfs list also lists the zpools and the disk images which are residing on them.
Question
How on earth can I get access to the disk images? I'd prefer to get them off of the external NVME and reformat it with ext4 or something, because I don't want to go through this nightmare again.
Hope someone knows what to do because Google, Gemini and me have no clue what to do.
My cluster consisting of 2 PVE hosts and a qdevice have been happily running for months, no issues. Last weekend I saw there were updates (it's my homelab, so pve-no-subscription updates), I ran them on both hosts, rebooted and everything seemed fine. Then I tried to use an USB ethernet adapter again purely for the management interface of host 1, that worked for a few hours, then the USB adapter lost it's connection (this issue occured earlier this year, but has been running fine for months before that). I undid the bridge configuration, rebooted host 1 and...
Issue
After the reboot host 1 disconnected an external USB NVME disk (zpool ext-1tb). I unplugged it, plugged it back in, disk LED became blue (good) then after a few seconds red (not good) then no light. Safe to say, VM's who had disk images on that zpool crapped out. In Disks - ZFS the status was suspended.
What I tried
I connected the USB disk to host 2, imported the zpool and voila, the disk appeared and was showing the raw disk images of a couple of VM's of host 1, proving there's nothing wrong with the disk or USB cable. However, I could not use qm disk import , it said the source file can't be found. zfs mount told me that pool ext-1tb is mounted at /ext-1tb. When I do ls -la in that folder, there's no folder or file showing, so makes sense that qm disk import didn't find the source. With zfs send | zfs recv I was able to copy/clone a disk image to local-zfs, which resides on the internal NVME disk of host 2. In the GUI, the disk image is indeed listed now, with it's size. But qm disk import still did not find any disk image. zfs list also lists the zpools and the disk images which are residing on them.
Question
How on earth can I get access to the disk images? I'd prefer to get them off of the external NVME and reformat it with ext4 or something, because I don't want to go through this nightmare again.
Hope someone knows what to do because Google, Gemini and me have no clue what to do.
