mxhuxley Posted June 13 Share Posted June 13 (edited) Hi all, I'm running unRAID 6.12.10, recently installed 2 new (to me) 16TB drives, precleared, with 1 added as parity (I went a while without a parity drive), and parity built, the other in the array. I have 3 very old, small hard drives that are at least on their way out, so the 2nd 16TB is already in the array. After an initial parity build, I replaced the 3TB bad drive with 1 of the 16TB's, which seems to have rebuilt successfully. I decided it was probably work starting with the drive with the most read errors on the parity build, and working my way down to the drive with the least read errors. I believe it also correlated with their capacity, if that matters. So 3TB had most, then 1TB, and 500GB with fewest read errors. I followed the instructions here: https://docs.unraid.net/legacy/FAQ/replacing-multiple-data-drives/ I'm running into issues with the remaining 2 drives, 1 more due to my ignorance than anything. The 3TB is in unassigned devices as I haven't physically disconnected it, in case diagnostics/testing/repair is advisable. It fails an extended SMART test for read errors, though the data at least appears to have rebuilt onto the larger drive. The 1TB (disk 6) seems to have copied over to the 16TB (disk 2), but I probably got the syntax wrong following the wiki, as it put the files in a 'disk 6' subfolder rather than the share. It also fails SMART extended with read errors. 16TB (disk 2) shows: /mnt/disk2/Seed <- The share that's on multiple disks, with some excluded /mnt/disk2/disk6/Seed <- The data copied via rsync from disk 6 to 2, but I need to move to the share on disk 2. I'm unsure on the syntax, as I tried a few GUI CA programs and mv in a terminal, but wound up with duplicated directory, so now structure is /mnt/disk2/Seed /mnt/disk2/Seed/Seed The bigger issue is that disk7 (500GB HDD) keeps giving issues with the same rsync command. Earlier I couldn't use ls as I was getting I/O errors, though a reboot fixed that. I'm running an extended SMART check in maintenance mode on it. I do have a bunch of new SATA cables on hand if swapping might prove helpful, but as I'm getting sector/read errors rather than CRC, I'm guessing it wouldn't do much? What's the best approach to get the data onto the new drive given all this? The subfolder issue is less of a concern, but I'd appreciate advice, and can make a separate thread for that if need be. Edited June 17 by mxhuxley Solved Quote Link to comment
mxhuxley Posted June 14 Author Share Posted June 14 (edited) So the 500GB (Disk 7) Extended Smart Test completed without errors. The dashboard tab SMART section shows a thumbs down, with 34 reallocated sectors and 10,465 reported uncorrect. I've attached the SMART report for Disk 7. For reference, the unassigned 3TB shows 10 current pending sectors and 10 offline uncorrectable, with the assigned 1TB (disk 6) at 46 online pending sectors and 6 offline uncorrectable. Both those drives immediately fail extended SMART tests for read errors, but seem to be functioning better than the 500GB for some reason? tower-smart-20240613-1808.zip Edited June 14 by mxhuxley Additional info on other disks Quote Link to comment
JorgeB Posted June 14 Share Posted June 14 7 hours ago, mxhuxley said: I decided it was probably work starting with the drive with the most read errors on the parity build, Do you mean there were read errors during the parity build? If yes, parity won't be 100% correct, if you didn't reboot yet post the diagnostics. Quote Link to comment
mxhuxley Posted June 14 Author Share Posted June 14 1 hour ago, JorgeB said: Do you mean there were read errors during the parity build? If yes, parity won't be 100% correct, if you didn't reboot yet post the diagnostics. There were read errors on the 3 drives both before I added the parity disc and during the initial parity build, as well as the disc rebuild from the pulled 3TB drive onto the new 16TB one. Unfortunately I've already rebooted. Quote Link to comment
JorgeB Posted June 14 Share Posted June 14 That means parity will likely be corrupt, you can try replacing the drives one by one, or just copy what you can from the old disks, that way at least you will now what is corrupt/missing. Quote Link to comment
mxhuxley Posted June 14 Author Share Posted June 14 12 hours ago, JorgeB said: That means parity will likely be corrupt, you can try replacing the drives one by one, or just copy what you can from the old disks, that way at least you will now what is corrupt/missing. Any advice since 2/3 drives are copied,but the 3rd one won't despite passing SMART? Should I do a filesystem repair? Quote Link to comment
JorgeB Posted June 14 Share Posted June 14 If the 3rd one is not mounting post the diags after a mount attempt. Quote Link to comment
mxhuxley Posted June 15 Author Share Posted June 15 (edited) 3 hours ago, JorgeB said: If the 3rd one is not mounting post the diags after a mount attempt. Fix Common Problems is stating 'Drive mounted read-only or completely full'. It has about 6GB free space of 500GB, so assuming it's mounted read-only. I can't imagine I need write access since I just want to pull the files off the drive? I'm not sure how to verify if it is in fact mounted read-only or rw. It's not showing as disabled on the unRAID 'Main' tab. I cannot access the drive via 'ls' in a terminal window due to I/O error. The drive shows up and I can 'cd' to it, but nothing more. Edited June 15 by mxhuxley Additional info Quote Link to comment
JorgeB Posted June 15 Share Posted June 15 Please post the diagnostics after mount. Quote Link to comment
mxhuxley Posted June 15 Author Share Posted June 15 6 hours ago, JorgeB said: Please post the diagnostics after mount. tower-diagnostics-20240615-1712.zip Quote Link to comment
Solution JorgeB Posted June 16 Solution Share Posted June 16 Check filesystem on disk7, run it without -n, also recommend running an extended SMART test, since SMART is showing some issues. Quote Link to comment
mxhuxley Posted June 16 Author Share Posted June 16 (edited) 18 hours ago, JorgeB said: Check filesystem on disk7, run it without -n, also recommend running an extended SMART test, since SMART is showing some issues. I ran without -n, got an error about empty logs, presumably because the drive was mounted read-only and thus can't write to log files. Running with -L seems to have fixed it, and I'm in process of copying drive contents via Rsync. Given that I already did an extended SMART test (zip linked above), and am pulling the drive from the array anyways, I'm assuming I don't need to run a second test? Edit: After finishing the filesystem check, I was able to pull the files via an 'mv' command. For getting rid of the duplicate subfolders from running rsync command, I enabled disk shares, moved the files out of the subfolders into the share folder (on the disk share, not the user share), deleted the empty subfolders, then disabled disk shares. I then did a new config, after physically removing the old/problematic drives. I didn't need to update included disks for my user shares, as the drives I removed (disks 3, 7, and 7) were either rebuilt (3TB to 16TB) or the last 2 disks in the array (disks 6/7), so the shares just removed them automatically. Fix common problems did complain they were still there, but following the instructions on the error, I just modified the share and hit 'apply' which fixed the issue after doing a rescan with the plugin. Thank you kindly Jorge for all the help! Edited June 17 by mxhuxley Closing Thread 1 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.