June 24, 20251 yr I recently stopped and started my array to made a change to an SMB setting and when the array came back up, my two-drive single mode cache pool won't mount. In the GUI the error shown is: "Unmountable: wrong or no filesystem".The unraid log shows:Jun 24 15:50:06 emhttpd: cache: btrfs verify devices Jun 24 15:50:06 emhttpd: /sbin/btrfs filesystem show af326f9e-a439-4fd0-b2ef-57c6d8da236c 2>&1 Jun 24 15:50:06 emhttpd: Label: none uuid: af326f9e-a439-4fd0-b2ef-57c6d8da236c Jun 24 15:50:06 emhttpd: #011Total devices 2 FS bytes used 502.52GiB Jun 24 15:50:06 emhttpd: #011devid 1 size 232.88GiB used 232.88GiB path /dev/sdc1 Jun 24 15:50:06 emhttpd: #011devid 2 size 931.51GiB used 335.00GiB path /dev/sdf Jun 24 15:50:06 emhttpd: cache: cannot import with misplaced devices Jun 24 15:50:06 emhttpd: shcmd (4542): rmdir /mnt/cache Jun 24 15:50:06 root: rmdir: failed to remove '/mnt/cache': Directory not empty Jun 24 15:50:06 emhttpd: shcmd (4542): exit status: 1 Jun 24 15:50:06 emhttpd: cache: mount error: wrong or no file systemYet the file system seems to be intact on both drives, and both are present. I can even mount the pool read only using mount -o rescue=all,ro /dev/sdc1 /tempOutput of btrfs fi show :Label: none uuid: af326f9e-a439-4fd0-b2ef-57c6d8da236c Total devices 2 FS bytes used 502.52GiB devid 1 size 232.88GiB used 232.88GiB path /dev/sdc1 devid 2 size 931.51GiB used 335.00GiB path /dev/sdfbtrfs check output for each drive with no errors:Checking filesystem on /dev/sdc1 UUID: af326f9e-a439-4fd0-b2ef-57c6d8da236c [1/8] checking log skipped (none written) [2/8] checking root items [3/8] checking extents [4/8] checking free space tree [5/8] checking fs roots [6/8] checking only csums items (without verifying data) [7/8] checking root refs [8/8] checking quota groups skipped (not enabled on this FS) found 539572510720 bytes used, no error found total csum bytes: 391171280 total tree bytes: 1112637440 total fs tree bytes: 493502464 total extent tree bytes: 156663808 btree space waste bytes: 189219902 file data blocks allocated: 16886634491904 referenced 482911248384Checking filesystem on /dev/sdf UUID: af326f9e-a439-4fd0-b2ef-57c6d8da236c [1/8] checking log skipped (none written) [2/8] checking root items [3/8] checking extents [4/8] checking free space tree [5/8] checking fs roots [6/8] checking only csums items (without verifying data) [7/8] checking root refs [8/8] checking quota groups skipped (not enabled on this FS) found 539572510720 bytes used, no error found total csum bytes: 391171280 total tree bytes: 1112637440 total fs tree bytes: 493502464 total extent tree bytes: 156663808 btree space waste bytes: 189219902 file data blocks allocated: 16886634491904 referenced 482911248384Output of fdisk -l /dev/sdc1Disk /dev/sdc1: 232.88 GiB, 250058301440 bytes, 488395120 sectors Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytesOutput of fdisk -l /dev/sdfDisk /dev/sdf: 931.51 GiB, 1000204886016 bytes, 1953525168 sectors Disk model: WD Blue SA510 2. Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytesdiagnostics-20250624-1557.zip Edited June 24, 20251 yr by thebluespirit
June 25, 20251 yr Community Expert Reboot and post new 10 hours ago, thebluespirit said:sdfUnraid should always use the partition, not the whole device, any idea how that happened?
June 25, 20251 yr Author Later today I will reboot and post new diagnostics. Though, the ones I sent were gathered after a reboot.7 hours ago, JorgeB said:any idea how that happened?Not really. The way it went down was I had the 250GB cache drive for years in a single drive pool, that's the Samsung (sdc1). Several weeks ago I got the 1TB WD drive (sdf). I expanded the pool slots to 2, added the new drive, and converted the pool to single mode so I could utilize all the space from both drives. The only thing I remember doing that deviated from default behavior was changing the pool from raid1 to single mode. Edited June 25, 20251 yr by thebluespirit Added detail
June 25, 20251 yr Community Expert 32 minutes ago, thebluespirit said:Though, the ones I sent were gathered after a reboot.In that case, no need for new ones, see if the pool mounts normally using the CLI, without rgw rescue option:mkdir /xmount -t btrfs /dev/sdc1 /xIf it mounts, post the output from btrfs fi usage -T /xIf the pool is a mirror and all looks good, the best way forward may be to offline the whole device and then add it back again and let Unraid rebuild it, it will create the missing partition, I can post more details after seeing that output.
June 25, 20251 yr Author Thanks Jorge, since I was able to mount it RO, I copied all the data off the cache drives and onto the array. I've had this cache pool for many years and through many UnRAID versions prior to adding the new disk, so I'm thinking it might be the best use of time to just start fresh with a new cache pool, then copy the data back. What I've done is:stop the array and erase the current cache pool, then removed it. I then added a new cache pool with 2 slots. I clicked on the first cache drive which in this case I will make the WD (sdf) and changed the file system to zfs striped (Which I believe should accomplish the same as btrfs single mode in the case of using all data on both drives)Changed the file system from auto to zfsChanged the allocation type to striped (2 devices)Started the arrayFormatted the unmountable cache disks using the format button at the button with the other array controlsOnce they finish formatting, in the GUI I am getting the same unmountable: wrong or no filesystem error. diagnostics-20250625-1041.zip
June 25, 20251 yr Author Solution I was finally able to get the new ZFS pool to start properly. There was a file that existed in /mnt/cache which was preventing the directory from being removed which was interrupting the formatting process. After removing the file I formatted the cache pool drives again and they were able to be mounted.I still don't know what happened to the original pool, but as I mentioned half of it was very old and half was very recently added. Not sure why the new drive wasn't partitioned by UnRAID, but that was a hint that something wasn't quite right.Thanks for your help @JorgeB Edited June 25, 20251 yr by thebluespirit
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.