Hi,

 

I installed the unraid upgrade from 7.0.1 to 7.1.2, after the reboot the cache pool shows as: "Unmountable: wrong or no file system".

 

Output of "zpool import":

  pool: cache
    id: 16823874084781934917
 state: ONLINE
status: Some supported features are not enabled on the pool.
        (Note that they may be intentionally disabled if the
        'compatibility' property is set.)
action: The pool can be imported using its name or numeric identifier, though
        some features will not be available without an explicit 'zpool upgrade'.
config:

        cache          ONLINE
          mirror-0     ONLINE
            nvme0n1p1  ONLINE
            nvme1n1p1  ONLINE

Output of "blkid":

/dev/sda1: LABEL_FATBOOT="UNRAID" LABEL="UNRAID" UUID="AB98-161C" BLOCK_SIZE="512" TYPE="vfat"
/dev/md1p1: UUID="4d34e813-c926-4d51-9f7e-1707eea4cbfa" BLOCK_SIZE="512" TYPE="xfs"
/dev/loop1: BLOCK_SIZE="131072" TYPE="squashfs"
/dev/nvme0n1p1: LABEL="cache" UUID="16823874084781934917" UUID_SUB="14463204487326902878" BLOCK_SIZE="4096" TYPE="zfs_member"
/dev/loop2: UUID="0abdf3dc-6e2c-42b9-b66a-0286e2f89501" UUID_SUB="4958cf0b-cda5-4a61-a06c-2f3f83a1c72e" BLOCK_SIZE="4096" TYPE="btrfs"
/dev/loop0: BLOCK_SIZE="131072" TYPE="squashfs"
/dev/nvme1n1p1: LABEL="cache" UUID="16823874084781934917" UUID_SUB="5643035414899293048" BLOCK_SIZE="4096" TYPE="zfs_member"

Output of "fdisk -l /dev/nvme0n1" and "fdisk -l /dev/nvme1n1"
 

root@DXP-2800:~# fdisk -l /dev/nvme0n1
Disk /dev/nvme0n1: 931.51 GiB, 1000204886016 bytes, 1953525168 sectors
Disk model: WDS100T3X0C-00SJG0                      
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 1953525167 1953523120 931.5G 83 Linux
root@DXP-2800:~# fdisk -l /dev/nvme1n1
Disk /dev/nvme1n1: 931.51 GiB, 1000204886016 bytes, 1953525168 sectors
Disk model: WD Red SN700 1000GB                     
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/nvme1n1p1       2048 1953525167 1953523120 931.5G 83 Linux

How can I get the cache back online? It's a pretty bad situation as the docker service files are stored on the cache and therefore docker does not run.

Thanks for your help.

See:

 

and

 

post the output of 

 

wipefs /dev/nvme1n1p1

39 minutes ago, nibu89 said:

the docker service files are stored on the cache

Unless you disabled Docker in Settings it has probably already created those on the array so you will need to clean that up. 

15 hours ago, MowMdown said:

post the output of 

 

wipefs /dev/nvme1n1p1

DEVICE    OFFSET       TYPE       UUID                 LABEL
nvme1n1p1 0x4000       zfs_member 16823874084781934917 cache
nvme1n1p1 0x44000      zfs_member 16823874084781934917 cache
nvme1n1p1 0xe8e0c04000 zfs_member 16823874084781934917 cache
nvme1n1p1 0xe8e0c44000 zfs_member 16823874084781934917 cache

 

Please post the diagnostics.

1 hour ago, JorgeB said:

Please post the diagnostics.

After a short search through syslog I found this:
 

May 11 15:04:17 DXP-2800 emhttpd: cache: import: misplaced device: /dev/nvme1n1p1
May 11 15:04:17 DXP-2800 emhttpd: errors: No known data errors
May 11 15:04:17 DXP-2800 emhttpd: cache: cannot import with misplaced devices

 

Also there seems to be no partion on /dev/nvme1n1 could this be a problem?

diagnostics are attached.

dxp-2800-diagnostics-20250512-1139.zip

Edited May 12, 20251 yr by nibu89

26 minutes ago, nibu89 said:

Also there seems to be no partion on /dev/nvme1n1 could this be a problem?

Yes, that the problem:

 

May 12 11:37:48 DXP-2800 emhttpd: /bin/lsblk -lnbo TYPE,PARTN,NAME,START,SIZE,PTTYPE,PARTTYPE /dev/nvme0n1 2>&1
May 12 11:37:48 DXP-2800 emhttpd: disk    nvme0n1         1000204886016 dos
May 12 11:37:48 DXP-2800 emhttpd: part  1 nvme0n1p1  2048 1000203837440 dos 0x83
May 12 11:37:48 DXP-2800 emhttpd: device nvme0n1 partition: nvme0n1p1 type: dos start: 2048 size: 976761560, code: 0x83 (4)
May 12 11:37:48 DXP-2800 emhttpd: import 30 pool device: (nvme0n1) WDS100T3X0C-00SJG0_21311P802617
May 12 11:37:48 DXP-2800 emhttpd: /bin/lsblk -lnbo TYPE,PARTN,NAME,START,SIZE,PTTYPE,PARTTYPE /dev/nvme1n1 2>&1
May 12 11:37:48 DXP-2800 emhttpd: disk    nvme1n1         1000204886016
May 12 11:37:48 DXP-2800 emhttpd: part  1 nvme1n1p1  2048 1000203837440  0x83
May 12 11:37:48 DXP-2800 emhttpd: device nvme1n1: no partitions

 

Note that for nvme1n1, the lsblk output is not showing the PTTYPE, it shows for the other device (dos)

 

May 12 11:37:48 DXP-2800 emhttpd: part  1 nvme0n1p1  2048 1000203837440 dos 0x83
May 12 11:37:48 DXP-2800 emhttpd: part  1 nvme1n1p1  2048 1000203837440  0x83

 

There's something wrong with that partition, and Unraid 7.1 relies on the lsblk output, so since some info is missing, it considers it an invalid partition.

 

Since the pool is a mirror, the easiest way to resolve this, would be to import the pool with just the good device, then wipe the other one and re-add it to the pool so that the partition gets correctly created, if you want to do that, type:

 

echo 1 > /sys/block/nvme1n1/device/device/remove

 

This will offline that device, then wait a couple of seconds and refresh the GUI, start the array with the remaining device, if all looks good, disable array auto start if enabled, reboot to bring the other device back online, wipe it with:

 

blkdiscard -f /dev/nvme1n1

 

Then add it back to the pool and start the array to resilver.

 

 

 

23 minutes ago, JorgeB said:

Since the pool is a mirror, the easiest way to resolve this, would be to import the pool with just the good device, then wipe the other one and re-add it to the pool so that the partition gets correctly created, if you want to do that, type:

 

echo 1 > /sys/block/nvme1n1/device/device/remove

 

This will offline that device, then wait a couple of seconds and refresh the GUI, start the array with the remaining device, if all looks good, disable array auto start if enabled, reboot to bring the other device back online, wipe it with:

 

blkdiscard -f /dev/nvme1n1

 

Then add it back to the pool and start the array to resilver.

Thanks everything is working again! Do you have any idea why or how that issue occurred?

10 minutes ago, nibu89 said:

Do you have any idea why or how that issue occurred?

Something happened when the partition was initially created, but unclear to me what, I've seen it a couple more times with other users, and it's always with NVMe devices, so possibly something in the kernel/Linux utils that sometimes doesn't create the partition 100% correct with NVMe devices.

As long as it only affects one device it's not that bad, but it would be nice to know if there is a way to avoid this.
Thanks again!