kenig Posted July 9, 2021 Share Posted July 9, 2021 (edited) Hi! I recently set up an unraid server and after some fiddling managed to get it working. Today I installed a new CPU fan and after starting up the array couldnt start all drives anymore. My setup: 12 TB parity (sdb) 12 TB data (sdc) 8TB data (sde) 256GB cache SSD (sda) the 12TB isn't mounting anymore. I get Unmountable: not mounted. After checking some forum posts I tried btrfs fi show and get both my 12TB drives with the same uuid: root@Tower:~# btrfs fi show /dev/sde1 WARNING: adding device /dev/sdc1 gen 2016 but found an existing device /dev/sdb1 gen 2032 ERROR: cannot scan /dev/sdc1: File exists WARNING: adding device /dev/md1 gen 2016 but found an existing device /dev/sdb1 gen 2032 ERROR: cannot scan /dev/md1: File exists Label: none uuid: 0b6913ce-61b3-4b16-9a03-3a82f0ef306b Total devices 1 FS bytes used 144.00KiB devid 1 size 7.28TiB used 2.02GiB path /dev/md2 root@Tower:~# btrfs fi show /dev/sda1 WARNING: adding device /dev/sdc1 gen 2016 but found an existing device /dev/sdb1 gen 2032 ERROR: cannot scan /dev/sdc1: File exists WARNING: adding device /dev/md1 gen 2016 but found an existing device /dev/sdb1 gen 2032 ERROR: cannot scan /dev/md1: File exists ERROR: not a valid btrfs filesystem: /boot root@Tower:~# btrfs fi show /dev/sdb1 WARNING: adding device /dev/sdc1 gen 2016 but found an existing device /dev/sdb1 gen 2032 ERROR: cannot scan /dev/sdc1: File exists WARNING: adding device /dev/md1 gen 2016 but found an existing device /dev/sdb1 gen 2032 ERROR: cannot scan /dev/md1: File exists Label: none uuid: 49d47325-9269-4f32-a2b3-6539ff1be045 Total devices 1 FS bytes used 5.31TiB devid 1 size 10.91TiB used 5.33TiB path /dev/sdb1 root@Tower:~# btrfs fi show /dev/sdc1 Label: none uuid: 49d47325-9269-4f32-a2b3-6539ff1be045 Total devices 1 FS bytes used 5.31TiB devid 1 size 10.91TiB used 5.33TiB path /dev/md1 The 8TB device is empty (3.7MB of 8TB used) I tried different SATA Ports. Unfortunately it made things more confusing as parity and data disk started switching places. Now I dont know which one is which. Can I use btrfstune -u /dev/sdc1 to get a new uuid for it? Is my data on that disk still available after that? Any help would be greatly appreciated! Edited July 9, 2021 by kenig Quote Link to comment
JorgeB Posted July 9, 2021 Share Posted July 9, 2021 Please post the diagnostics: Tools -> Diagnostics (after array start) Quote Link to comment
kenig Posted July 9, 2021 Author Share Posted July 9, 2021 Here you go tower-diagnostics-20210709-1036.zip Quote Link to comment
JorgeB Posted July 9, 2021 Share Posted July 9, 2021 Sometimes with few data disks (or an odd number of data disks), parity can appear to have a valid filesystem, this can happen with any filesystem but it's particularly a problem with btrfs since it scans for all filesystems at boot, and then it gets confused, solution here would be to use parity2 slot instead, the way it's calculated would make it impossible, or at least very unlikely, that a valid filesystem could be detected there, you'd need to unassign parity, wipe it with: wipefs -a /dev/sdb1 Reboot, then assign it as parity2 and let it sync. Quote Link to comment
kenig Posted July 9, 2021 Author Share Posted July 9, 2021 (edited) Thanks! Is there a way to tell which disk has been parity and which one was data? The problematic disc is sdc (12TB data drive) at the moment. I fear loosing data if I wipe parity and the drive has failed too. I tried mounting both with "Unassigned Drives" Plugin but neither could be mounted. Jul 9 11:44:39 Tower unassigned.devices: Mount drive command: /sbin/mount -t btrfs -o rw,auto,async,noatime,nodiratime '/dev/sdc1' '/mnt/disks/WDC_WD120EDAZ-11F3RA0_5PJHXG4C_2' Jul 9 11:44:39 Tower unassigned.devices: Mount of '/dev/sdc1' failed: 'mount: /mnt/disks/WDC_WD120EDAZ-11F3RA0_5PJHXG4C_2: mount(2) system call failed: File exists. ' Edited July 9, 2021 by kenig more info Quote Link to comment
JorgeB Posted July 9, 2021 Share Posted July 9, 2021 It's normal not mounting with UD because of the same reason, If you want to make sure all is well before wiping parity just disconnect parity from the server and after booting disk1 should now mount. Quote Link to comment
kenig Posted July 9, 2021 Author Share Posted July 9, 2021 Ok, will do that and then wipe parity if data is ok Quote Link to comment
kenig Posted July 9, 2021 Author Share Posted July 9, 2021 Data is ok. Will have to wait to rebuild parity until tonight. Need the array for work atm. Thanks a lot for that perfect support! 1 Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.