jehan Posted March 24 Share Posted March 24 (edited) Hi, In my Unraid array, I have five ZFS single-disk stripes, labeled Disk 1 through Disk 5, along with one parity drive. The five ZFS disks have been formatted through the Unraid GUI. Since I encountered an issue with Disk 2, I did and completed a parity check, stopped the array, removed the disk, and began clearing it. After a few minutes, I canceled the clearing process, reinserted the disk into the array, and restarted it, assuming Unraid would recognize Disk 2 as failed and initiate a rebuild process. Upon restarting the array, Disk 2 was flagged as "Unmountable: Unsupported partition layout," which was somewhat expected. However, what surprised me was that Disk 5 was flagged as "Unmountable: Unsupported or no filesystem," and Unraid suggests rebuilding Disk 5, which I accepted. Unfortunately, Disk 5 remains unmountable even after the rebuild has finished. Additionally, what makes me wonder is that the "zpool import" command shows that disk sdi (Disk 5) has the pool name "disk4." root@unraid:~# zpool import pool: disk4 id: 7004989206944992445 state: UNAVAIL status: One or more devices contains corrupted data. action: The pool cannot be imported due to damaged devices or data. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-5E config: disk4 UNAVAIL insufficient replicas sdi1 UNAVAIL invalid label syslog: Mar 23 19:42:20 unraid emhttpd: mounting /mnt/disk4 Mar 23 19:42:20 unraid emhttpd: shcmd (58): mkdir -p /mnt/disk4 Mar 23 19:42:20 unraid emhttpd: /usr/sbin/zpool import -f -d /dev/md4p1 2>&1 Mar 23 19:42:23 unraid emhttpd: pool: disk4 Mar 23 19:42:23 unraid emhttpd: id: 12000237853854133759 Mar 23 19:42:23 unraid emhttpd: shcmd (59): /usr/sbin/zpool import -f -N -o autoexpand=on -d /dev/md4p1 12000237853854133759 disk4 Mar 23 19:42:27 unraid emhttpd: shcmd (60): /usr/sbin/zpool online -e disk4 /dev/md4p1 Mar 23 19:42:27 unraid emhttpd: /usr/sbin/zpool status -PL disk4 2>&1 Mar 23 19:42:27 unraid emhttpd: pool: disk4 Mar 23 19:42:27 unraid emhttpd: state: ONLINE Mar 23 19:42:27 unraid emhttpd: config: Mar 23 19:42:27 unraid emhttpd: NAME STATE READ WRITE CKSUM Mar 23 19:42:27 unraid emhttpd: disk4 ONLINE 0 0 0 Mar 23 19:42:27 unraid emhttpd: /dev/md4p1 ONLINE 0 0 0 Mar 23 19:42:27 unraid emhttpd: errors: No known data errors Mar 23 19:42:27 unraid emhttpd: shcmd (61): /usr/sbin/zfs set mountpoint=/mnt/disk4 disk4 Mar 23 19:42:28 unraid emhttpd: shcmd (62): /usr/sbin/zfs set atime=off disk4 Mar 23 19:42:28 unraid emhttpd: shcmd (63): /usr/sbin/zfs mount disk4 Mar 23 19:42:28 unraid emhttpd: shcmd (64): /usr/sbin/zpool set autotrim=off disk4 Mar 23 19:42:29 unraid emhttpd: shcmd (65): /usr/sbin/zfs set compression=off disk4 Mar 23 19:42:29 unraid emhttpd: mounting /mnt/disk5 Mar 23 19:42:29 unraid emhttpd: shcmd (66): mkdir -p /mnt/disk5 Mar 23 19:42:29 unraid emhttpd: /usr/sbin/zpool import -f -d /dev/md5p1 2>&1 Mar 23 19:42:32 unraid emhttpd: pool: disk4 Mar 23 19:42:32 unraid emhttpd: id: 7004989206944992445 Mar 23 19:42:32 unraid emhttpd: shcmd (68): /usr/sbin/zpool import -f -N -o autoexpand=on -d /dev/md5p1 7004989206944992445 disk5 Mar 23 19:42:35 unraid root: cannot import 'disk4' as 'disk5': one or more devices is currently unavailable Mar 23 19:42:35 unraid emhttpd: shcmd (68): exit status: 1 Mar 23 19:42:35 unraid emhttpd: shcmd (69): /usr/sbin/zpool online -e disk5 /dev/md5p1 Mar 23 19:42:35 unraid root: cannot open 'disk5': no such pool Mar 23 19:42:35 unraid emhttpd: shcmd (69): exit status: 1 Mar 23 19:42:35 unraid emhttpd: disk5: import error Mar 23 19:42:35 unraid emhttpd: /mnt/disk5 mount error: Unsupported or no file system Is there anything I can try to restore the ZFS filesystem on Disk 5 so that Unraid can mount it? Thank you for your help unraid-diagnostics-20240324-1349.zip Edited March 24 by jehan Quote Link to comment
JorgeB Posted March 24 Share Posted March 24 3 hours ago, jehan said: Since I encountered an issue with Disk 2, I did and completed a parity check, stopped the array, removed the disk, and began clearing it. After a few minutes, I canceled the clearing process, reinserted the disk into the array, and restarted it, assuming Unraid would recognize Disk 2 as failed and initiate a rebuild process. Not sure I follow, was disk2 disabled at any point, or did you just unassigned it, cleared and re-assign it without it being disabled? That would make parity invalid. And was disk5 disabled or just unmountable? Quote Link to comment
jehan Posted March 24 Author Share Posted March 24 What I did was: stop the array, remove Disk 2, start the array - Disk 2 was emulated. Than I started to clear Disk 2. Than I stopped the array and plugged Disk 2 back to the array and started the array. At this point Disk 5 became unavailable. I haven't disabled Disk 5 at any point actively. I hope I answered your question. Quote Link to comment
JorgeB Posted March 24 Share Posted March 24 2 hours ago, jehan said: Than I stopped the array and plugged Disk 2 back to the array and started the array. That would begin a rebuild, it could never show 8 hours ago, jehan said: "Unmountable: Unsupported partition layout," Something is missing, and the diags don't cover before that, so we can't see what happened. I do see that besides the other issues, zfs detected data corruption in disks 1 and 3, suggesting you have a serious hardware issue: Mar 23 19:42:11 unraid emhttpd: status: One or more devices has experienced an error resulting in data Mar 23 19:42:11 unraid emhttpd: corruption. Applications may be affected. Start by running memtest. Quote Link to comment
jehan Posted March 25 Author Share Posted March 25 Yes, I'm aware of the detected data corruption. Previously, I had some problems with newly attached hardware which forced me to hard-reset the system several times, causing the data corruption. However, this is not a significant issue since I have the corrupted data backed up. I previously conducted a memtest, which showed no problems with the memory. Upon plugging Disk 2 back into the array, Disk 5 became unavailable after I rebooted the Unraid machine. Consequently, I had two disks that could not be mounted. Surprisingly (for me), Unraid offered me the option to rebuild Disk 5, which I accepted. What do you think? Do I have any chance of rescuing the data on Disk 5? Quote Link to comment
JorgeB Posted March 25 Share Posted March 25 3 minutes ago, jehan said: Previously, I had some problems with newly attached hardware which forced me to hard-reset the system several times, causing the data corruption. Hard reset should never cause data corruption, they can cause some data to be lost, if you were in the middle of writing to the pool, but in those cases the data should be complete or incomplete, never corrupt. Also note that using USB for the array/pool is strongly discouraged, it can cause several issues, possibly related to at least some you are seeing: Mar 23 19:42:35 unraid root: cannot import 'disk4' as 'disk5': one or more devices is currently unavailable This suggest disk5 was disk4 at some point in the past, unfortunately the diags don't cover what happened before, post the output of zpool import Quote Link to comment
jehan Posted March 25 Author Share Posted March 25 8 minutes ago, JorgeB said: Hard reset should never cause data corruption, they can cause some data to be lost, if you were in the middle of writing to the pool, but in those cases the data should be complete or incomplete, never corrupt. Okay - good to know. 9 minutes ago, JorgeB said: Also note that using USB for the array/pool is strongly discouraged, it can cause several issues, possibly related to at least some you are seeing Yes - I'm a bit limited on my side here. zpool import root@unraid:~# zpool import pool: disk4 id: 7004989206944992445 state: UNAVAIL status: One or more devices contains corrupted data. action: The pool cannot be imported due to damaged devices or data. see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-5E config: disk4 UNAVAIL insufficient replicas sdh1 UNAVAIL invalid label root@unraid:~# Quote Link to comment
JorgeB Posted March 25 Share Posted March 25 Looks like another corrupt pool, though in this case the label/metadata is corrupt, so it cannot even be imported, and like mentioned this pool was disk4 at some point in the past. Unfortunately, and AFAIK, there's nothing than can be done for this error other than destroying the pool and restoring from a backup. Quote Link to comment
jehan Posted March 25 Author Share Posted March 25 ok Thank you very much for all your efforts. 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.