March 6, 20251 yr Good day, Since yesterday (after a hard shutdown i had to do because the server was no longer reachable), i have gotten the above error for my One Cache Drive i use since Dec 2023. I have done the following research and fix steps already (following loosely this thread here). I have also tried: Restart, stop array, reassign mounts, start array, Restore from backup earlier that day (couldnt find an older version). The following (hopefulyl relevant) commands return the following: btrfs fi show Label: none uuid: 46f2df36-e9e3-4e60-b618-fa56936b6a5a Total devices 1 FS bytes used 340.00KiB devid 1 size 20.00GiB used 536.00MiB path /dev/loop2 Label: none uuid: cef1ec31-8a35-4148-93d3-2c1ddde40ff0 Total devices 1 FS bytes used 783.53GiB devid 1 size 1.82TiB used 851.07GiB path /dev/nvme0n1p1 So apparently the device is not completely corrupt. II tried creating a read-only mount next and creating a backup, as recommended here mount -o usebackuproot,ro /dev/nvme0n1p1 /temp this showed me no errors, it just took a while. Browsing in /temp also showed my old data structure As far as i am aware, this is just a shallow copy in the RAM so i wanted to do an actual backup in case something goes wrong du -sh /temp ls /temp | wc -l This also completed after some time Afterwards i got recommendet to check if the backup checksums were matching: btrfs restore -l /dev/nvme0n1p1 | wc -l find /temp -type f | wc -l which they were not (first command = 314, second command = 0). i Then cautiously went back to the original helpthread, and tried to go from there > fdisk -l /dev/nvme0n1p1 Disk /dev/nvme0n1p1: 1.82 TiB, 2000397885440 bytes, 3907027120 sectors Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes whcih is a different result than the user in that post was getting. And the fix that was proposed typing in sfdisk /dev/nvme0n1p1 with 2048 gives me Welcome to sfdisk (util-linux 2.38.1). Changes will remain in memory only, until you decide to write them. Be careful before using the write command. Checking that no-one is using this disk right now ... OK The device contains 'btrfs' signature and it may be removed by a write command. See sfdisk(8) man page and --wipe option for more details. Disk /dev/nvme0n1p1: 1.82 TiB, 2000397885440 bytes, 3907027120 sectors Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes sfdisk is going to create a new 'dos' disk label. Use 'label: <name>' before you define a first partition to override the default. Type 'help' to get more information. >>> 2048 Created a new DOS disklabel with disk identifier 0x31d59598. The device contains 'btrfs' signature and it may be removed by a write command. See sfdisk(8) man page and --wipe option for more details. Created a new partition 1 of type 'Linux' and of size 1.8 TiB. /dev/nvme0n1p1p1 : 2048 3907027119 (1.8T) Linux /dev/nvme0n1p1p2: Which is a different thing that the user in the support thread i referenced got. So i CTRL+C it, and tried with 64 Welcome to sfdisk (util-linux 2.38.1). Changes will remain in memory only, until you decide to write them. Be careful before using the write command. Checking that no-one is using this disk right now ... OK The device contains 'btrfs' signature and it may be removed by a write command. See sfdisk(8) man page and --wipe option for more details. Disk /dev/nvme0n1p1: 1.82 TiB, 2000397885440 bytes, 3907027120 sectors Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes sfdisk is going to create a new 'dos' disk label. Use 'label: <name>' before you define a first partition to override the default. Type 'help' to get more information. >>> 64 Created a new DOS disklabel with disk identifier 0xb8388533. The device contains 'btrfs' signature and it may be removed by a write command. See sfdisk(8) man page and --wipe option for more details. Created a new partition 1 of type 'Linux' and of size 1.8 TiB. /dev/nvme0n1p1p1 : 64 3907027119 (1.8T) Linux /dev/nvme0n1p1p2: Which is also different. At that point i was too afraid to continue in fear of messing something up with my amateur workings. I also looked into it (after unmounting) with btrfs check /dev/nvme0n1p1 Which gave me a pretty long dump of text i dont want to paste here, i attached it. But all in all there were multiple errors found. So i am left at this point, and would love to have some professional opinions on what my next steps should be. I didnt want to try "restore" or simliar because in the "FAQ for Unraid" it gets called a lsat resort, able to mess stuff up. Thanks a lot for your time! btrfs check.txt
March 6, 20251 yr Community Expert 1 hour ago, APM said: fdisk -l /dev/nvme0n1p1 This should not include the partition, so fdisk -l /dev/nvme0n1 Also, please post the diagnostics after array start, so that it shows why the pool doesn't mount
March 6, 20251 yr Author 1 hour ago, JorgeB said: This should not include the partition, so fdisk -l /dev/nvme0n1 Also, please post the diagnostics after array start, so that it shows why the pool doesn't mount Thank you for your answer, yeah the "p1" being part of it was indeed an oversight. fdisk -l /dev/nvme0n1 gives Disk /dev/nvme0n1: 1.82 TiB, 2000398934016 bytes, 3907029168 sectors Disk model: Samsung SSD 970 EVO Plus 2TB Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disklabel type: dos Disk identifier: 0x00000000 Device Boot Start End Sectors Size Id Type /dev/nvme0n1p1 2048 3907029167 3907027120 1.8T 83 Linux Which lines up again with the data from this problem ( After starting the array, i downloaded the diagnostics like you mentioned. diagnostics-20250306-1414.zip
March 6, 20251 yr Community Expert The partition looks fine, this is the problem: Mar 6 14:06:44 Tower kernel: BTRFS error (device nvme0n1p1): parent transid verify failed on logical 553746432 mirror 1 wanted 9835936375106488927 found 1100383 Mar 6 14:06:44 Tower kernel: BTRFS error (device nvme0n1p1): parent transid verify failed on logical 553746432 mirror 2 wanted 9835936375106488927 found 1100383 This is a very suspect number, when there's for example write barrier issue, this can sometimes be off by a few decimals, so something really wrong happened there, and both metadata copies are corrupt in the same way, this error is fatal with btrfs, it cannot be repaired, looks by your description that you already backed up the data? If not, you can also try btrfs restore.
March 6, 20251 yr Author Thanks, I just did a shallow copy to /temp (sorry not that deep in bash functionality), but i believe thats just some parts loaded in ram. Essentially i just used that to check if my files are still there (it still had my shares as folders etc.). I dont think there is data on there thats necessary to be kept, so i guess Reformatting it is? Can i just use the Unraid GUI feature for that and create a new partition? Or is there a way i can load off everythign currently stored on there to some place on my two other disks that i use as array? I am not even sure whcih parts of my files are located on the SSD and which arent, the structure of Unraid working in shares makes this a bit hard for me to understand sorry. Thanks for your help
March 6, 20251 yr Community Expert If there's no data that you need in the pool, just start the array and reformat it.
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.