Sorry you ran into this problem, looks like one of you drives(sde) is missing from the system. We’ll need to do some system/BTRFS level troubleshooting to find out more. Can you provide the output of the following commands? ( you need to run them as root user on your Rockstor system)
I see the problem. udevadm recognizes /dev/sde but blkid does not. btrfs uses blkid for device scanning, so it doesn’t see your sde(and others with the same problem) and hence you have devices missing. Try the following commands:
btrfs device scan /dev/sde
btrfs fi show
I am wondering if specifying the device explicitly might help the scan. Please provide output of the commands and any extra information you think is relevant. A screenshot of your pools table can also be useful.
Also this is not related to Rockstor update and a BTRFS troubleshooting issue. So, I’ve updated the topic headline.
It seems like Rockstor doesn’t think sde is part of any Pool, but it does see it. You can add it from the Web-UI. Take a look at this documentation on how to add a disk via pool resize. There’s also a screencast in there that shows you exactly how. Add sde to storage1 and then let’s see if the output of btrfs fi show changes.
Besides that, it’s very mysterious how the device just dropped. If you can, please send me the logs located at /opt/rockstor/var/log. You can just zip it and e-mail to firstname.lastname@example.org.
No luck. Every time I reboot the NAS, drive sde disappears. I add it again and rebalance takes a day.
I tried to change the RAID from 10 to 1 and rebalance so I can mount the drive and get to my data but that also throws an error.
Error running a command. cmd = [’/bin/mount’, u’/dev/disk/by-label/storage1’, u’/mnt2/storage1’, ‘-o’, u’,compress=no’]. rc = 32. stdout = [’’]. stderr = [‘mount: wrong fs type, bad option, bad superblock on /dev/sdb,’, ’ missing codepage or helper program, or other error’, ‘’, ’ In some cases useful info is found in syslog - try’, ’ dmesg | tail or so.’, ‘’]
bummer. So everytime you reboot, btrfs fi show command reports missing device(supposedly sde)? This remains the same even after you explicitly run btrfs device scan? But you are able to add the device via pool resize and rebalance?
I wish I could reproduce this behavior. I brought it up on #btrfs IRC and no one seemed to have seen this before. I suggest you report this to btrfs mailing list. They will help you troubleshoot more and I’ll help anyway I can.
I had to resize/rebalance the pool from a RAID10 to a RAID1
Then delete missing drives (sde)
Reboot, which mounted the pool and shares as RAID1
Then I had to delete the superblocks on sde
Then add sde back to the converted RAID1 pool
I think I am going to leave it as a RAID1 and not mess around anymore.
I dont know if it is a similar scenario, but I experienced similar problems when I was testing Rockstor the first time.
I had made a pool on a single disk, that I had given a name (TestPool).
Seing that RockStor installed and worked well, I took 4 other disks, reinstalled, and made a Raid5 also named TestPool. This worked well, so I decided to add the first disk, from the first install, to this pool. To my surprise RockStor took that drive and added it to the RAID5 without any intervention, allthough it was not part of it. I at first thought this was by design (which it wasn’t), but later I found out it had mixed the single disk pool and the RAID5 pool, probably because they had the same name.
This eventually messed up both pools completely, and I had to rebuild.
As this was part of a test, I wasn’t that worried, and have just been carefull to remove existing partitions from new disks before adding them.
But perhaps what is being seen is the same sort of behaviour?