I’m having a similar situation to this one at the moment:
I extended a single pool to be a raid1 by adding another disk. Did a balance, which somehow started, but didn’t register in the Rockstor UI. After the balance was finished everything was fine, but once I rebooted the pools where messed up in Rockstor.
It looks like the default rockstor boot for the installation disappeared, and the disk that is used when installing rockstor now mounts under /mnt2/slow (which is the name of the pool that disappeared. Rockstor still boots up, and the other pool I have is still shows up. The data of the old pool is still there too (thank god, gave me quite a fright at first).
I think I’m going to re-install the whole thing, and re-import my pools. Except there is an easier way to get back my old disk layout? I
I’m on the stable channel with a subscription, for what it’s worth.
Hm, it looks like the 2 disks that are part of the raid 1 that disappeared are ‘merged’ with the rockstor install disk in the web-UI (showing up as single with 3 disks). I don’t dare to remove the disk from there, don’t want to kick of any btrfs command that could destroy my data.
thanks. Unfortunately I needed my NAS up, so I had to reinstall everything. I figured I’d post this here on the off-chance somebody would respond within the hour, but also to record the issue in some way, in case somebody else comes across it.
The btrfs commands showed nothing unusual, the raid was a proper btrfs raid (Data & metadata was RAID1, with the 2 proper disks-devs listed), and the rockstor partition was a single. It looks like only the rockstor code somehow mixed things up, not sure. The UI showed only the one pool with the 3 disks in question, and as a single, using the non-rockstor pool name. Only the subvolumes of the rockstor partition were mounted under /mnt2.
Anyway, after the reinstall the raid and it’s subvolumes were properly recognized and imported, so all good again.