Hi guys,

normally I do not ask for help and prefer to read and troubleshoot by myselft but I am at a point where all related topics did not lead me to a solution: after a couple of (hardware related) forced reboots of my server the cache pool came up "Unmountable: Unsupported or no file system". Attached find my diagnostics. Help is much appreciated!

server-diagnostics-20240113-1247.zip

No valid filesystem is being detected, do you know if the pool was btrfs or zfs? Also post the output of:

blkid

and

fdisk -l /dev/nvme0n1

 

Quote

No valid filesystem is being detected, do you know if the pool was btrfs or zfs? Also post the output of:

blkid

 

root@SERVER:~# blkid
/dev/sdb1: LABEL_FATBOOT="UNRAID" LABEL="UNRAID" UUID="2732-64F5" BLOCK_SIZE="512" TYPE="vfat"
/dev/sda1: UUID="7c22b33e-a8c5-4fe7-9852-d3fa69fd4aed" UUID_SUB="2a81485a-1faf-4be3-bcb0-bd636a5943b4" BLOCK_SIZE="4096" TYPE="btrfs"
/dev/loop1: TYPE="squashfs"
/dev/md1p1: UUID="7c22b33e-a8c5-4fe7-9852-d3fa69fd4aed" UUID_SUB="2a81485a-1faf-4be3-bcb0-bd636a5943b4" BLOCK_SIZE="4096" TYPE="btrfs"
/dev/loop0: TYPE="squashfs"
/dev/sdc1: UUID="7c22b33e-a8c5-4fe7-9852-d3fa69fd4aed" UUID_SUB="2a81485a-1faf-4be3-bcb0-bd636a5943b4" BLOCK_SIZE="4096" TYPE="btrfs"

 

Quote

and

fdisk -l /dev/nvme0n1

 

root@SERVER:~# fdisk -l /dev/nvme0n1
Disk /dev/nvme0n1: 931.51 GiB, 1000204886016 bytes, 1953525168 sectors
Disk model: CT1000P3SSD8                            
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes

 

I think chache was btrfs before but I am not 100% sure. I already tried: Stop Array -> Remove Cache Devices -> Start Array -> Stop Array -> Re-Add Cache Drives with Filesystem "auto". I also tried to manually chanche the FS to zfs and btrfs resulting in the same error.

There is no partition on the NVMe device, suggesting the device was wiped, the other one is likely the same, a forced reboot should never cause that, post the output of:

 

sfdisk /dev/nvme0n1

then type

2048

and hit return and post a screenshot of the results

 

Type N to KEEP the signature, hit return, then type

write

and hit return again, do the same for the other nvme device (nvme1n1) and then post the output of

btrfs fi show

 

P.S. I may only be able to reply tomorrow, so you can also try re-starting the array to see if the pool mounts, it won't damaged anything if it doesn't, if you just re-start it as is, and post new diags if it still doesn't mount

 

 

Just did the steps and this is the result for btrfs fi show:

 

root@SERVER:~# btrfs fi show
Label: none  uuid: 7c22b33e-a8c5-4fe7-9852-d3fa69fd4aed
        Total devices 1 FS bytes used 787.69GiB
        devid    1 size 1.82TiB used 883.02GiB path /dev/md1p1

Label: none  uuid: 912de6db-b116-4e8f-b947-5cde535b860c
        Total devices 2 FS bytes used 276.66GiB
        devid    1 size 931.51GiB used 489.01GiB path /dev/nvme0n1p1
        devid    2 size 931.51GiB used 489.01GiB path /dev/nvme1n1p1

I will re-start the array now

Ok, sadly no luck. I restarted the host just to be sure but still the same error. Attached find a new diagnostic.

Thank you for your time and effort, I really apreciate the help!

server-diagnostics-20240113-1955.zip

The filesystem is now detected but there's an issue with the log tree, if that is the only problem this may help:

 

btrfs rescue zero-log /dev/nvme0n1p1

 

Then re-start the array.

root@SERVER:~# btrfs rescue zero-log /dev/nvme0n1p1
Clearing log on /dev/nvme0n1p1, previous log_root 1167934586880, level 0

After restarting the array everything is working again. Thank you very, very much @JorgeB