January 31, 20251 yr I've been having on and off issues with my array for some time and things kinda came to a head today. I have been trying to get a parity check to complete so that I can do some playing with my disks and get things fixed but my disk 4 kept hard resetting causing the parity check to take forever. Then for whatever reason my disk 1 threw an error and went emulated so I ended up pulling the drive and replacing it, during the sync the disk 4 was causing the same issues and the sync was going to take several years. So I stopped the sync and the array, messed with the power cables of the drives as that is what I think was causing the issue, when I brought up the array multiple drives failed to mount, ironically not drive 4 but rather drive 1 (the one that was emulated) 3 and 5. Went to maintenance mode and scanned them, when the drives went back up 1 and 3 remained unmountable but 5 returned good. Unraid then decided to run the parity check and quickly disk 3 thew an error and went emulated as well.... I only have one parity drive so this is obviously not good. So now I have disk 1 and 3 unmountable, but the parity check ran and completed (quite quickly). I am currently assuming disk 3 is actually ok just with a read error during all the chaos, I also have the original disk 1 (currently not connected) which I believe is more or less correct but I haven't mounted it anywhere to check. At the exact second I have the array mounted in maintenance mode. Any clue on what my next steps should be to try to get these drives back up with the least amount of missing data. oden-diagnostics-20250131-0019.zip
January 31, 20251 yr Community Expert Check/replace cables for disks 3 and 5 and post new diags after array start.
January 31, 20251 yr Author Array is all sorts of unhappy As it sits, Disk 1 should be empty (mostly) as it was the drive I was syncing at the beginning of this but I have the original disk that I hope is readable. and disk 3 was good before all this chaos. If I could get 3 remounted with its content, I could format disk 1 (as since parity is written over there would be no rebuilding that directly) or remount the old disk 1 in the same way (if possible) then get parity up to snuff I would be good (all though I would most likely need some fixing of dockers and internal content but at least I would have the contents back). oden-diagnostics-20250131-0933.zip
January 31, 20251 yr Community Expert Now there are issues with parity, are you using power splitters? Could also be a bad PSU.
January 31, 20251 yr Author Yeah I have power splitters, I think that what was giving me the original intermittent issue with disk 4, so at the start of this I moved the power around to hopefully resolve the issue, which clearly did not work out. I have some new splitters on the way to at least swap out the one that I think was causing the issue.
January 31, 20251 yr Community Expert 1 hour ago, armagedon41 said: I have some new splitters on the way to at least swap out the one that I think was causing the issue. If using SATA->sata splitters you should not split more than 2 way to avoid issues. With Molex->sata you can normally get away with 4 way
January 31, 20251 yr Author Doing Molex to sata splitting, only have 2 2way splitters on different branches so it should be fine but I suspect the one has gotten loose over time so I will try and replace it just in case.
February 1, 20251 yr Author Ok changed the splitters and more or less in the same state At least no parity errors this time oden-diagnostics-20250131-2345.zip
February 1, 20251 yr Community Expert No more errors for now, check filesystem on both disk1 and the emulated disk3, run it without -n.
February 1, 20251 yr I had a similar issue with the usual (and good) advice of "check power and sata cables". I had 5 disks on one lead from the PSU. I added another PSU lead and shared the 5 as 3 and 2 disks. I also KEPT the DISKS SPINNING. No spindown and the array has been stable for about 3 weeks now.
February 2, 20251 yr Author Scan on disk 3 is showing Phase 1 - find and verify superblock... couldn't verify primary superblock - not enough secondary superblocks with matching geometry !!! attempting to find secondary superblock... .found candidate secondary superblock... unable to verify superblock, continuing... ................................................................ unable to verify superblock, continuing... Disk 1 is looking good, mostly empty but expected. Still just that disk 3 oden-diagnostics-20250202-1138.zip
February 2, 20251 yr Community Expert 32 minutes ago, armagedon41 said: Scan on disk 3 is showing Are you doing this on emulated disk3 from the webUI? Easy to get the command wrong, webUI will get it right.
February 2, 20251 yr Author yeah, left it overnight and the "...." just kept getting longer then finally jumped to the next line with "continuing". I had to cancel it. I tried it a couple more times with the same result.
February 2, 20251 yr Author May have been more power issues, I mounted (read only) my old disk 1 to start to copy the files back to the new disk 1 using mc and it would start to copy then die a few seconds in and the parity would have tens of thousands of errors and the new disk 1 would even unmount on one attempt. So I swapped out the power supply completely and the copy is working this time. Once this copy is complete I will drop to maintenance mode and scan disk 3, in the meantime if any one has any suggestion if the scan does the same thing it would be appreciated.
February 2, 20251 yr Community Expert On 1/31/2025 at 2:48 AM, armagedon41 said: I am currently assuming disk 3 is actually ok just with a read error during all the chaos SMART for physical disk3 looks OK. You could try to unassign it and see if it will mount as an Unassigned Device. If not, we could try check filesystem on it also. If we can get it mountable, we could New Config it back into the array and rebuild parity.
February 2, 20251 yr Author mc is moving that data over slowly, 75 MB/s so have about a day and a half, might try and stop and unassign before its all done, but I will update either after the copy or if I plan to live on the edge 😃
February 4, 20251 yr Author I was able to mount it in unassigned devices. not sure what the next step would be to get it back in to Disk 3 without it clearing it. On 2/2/2025 at 4:45 PM, trurl said: we could New Config it back into the array and rebuild parity.
February 4, 20251 yr Author I am assuming that I need to just put back the original disk in to disk 3 (now showing as new) then run the New Config with preserve current assignments set to all? (this is where I an not sure) And then assuming the array does not auto start, start it back up and let parity check run then everything should be back up with all the data in on all the disks.
February 5, 20251 yr Community Expert Solution Preserve all, then you have to check the Yes, I want to do this box, then it will enable the Apply button which you have to push. That will get you to the disk assignment page with the array stopped. You can make any changes needed, such as reassigning disk3. Main thing is to not assign a data disk to any parity slot or it will be overwritten by parity. Then start the array to begin parity rebuild (don't check the Parity Valid box).
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.