May 30, 20251 yr Hello, Looking for some guidance on the best process to recover. 24 Bay Enclosure, 16 total disks, single parity, using the Supermicro X9DR3-F.I had a single disk (disk 5) drop out as invalid and was emulated but the drive passes SMART test. I only see a few pending sectors for it . After a power failure I lost another drive entirely and isn't detected on any system. (Disk 8). Data on either of the drives is not critical but would preferable to not lose it.Ideal scenario, I force drive 5 back into the array and emulate Disk 8 for a rebuild.Second best would be to recover the array with whatever data is currently on all of the drives aside from the failed #8.Worst case scenario, Neither 5 or 8 have anything irreplaceable and I would just rebuild the array without those 2 drives.I have 2 new parity drives on order to upgrade the size and replace the single parity. For options 2 and 3 I would do the rebuild in a few days and add/replace the parity disks at the same time, in case that changes the advice.Diagnostics are attached.Thanks! tower-diagnostics-20250530-1447.zip
May 31, 20251 yr Community Expert How long did it pass from disk5 getting disabled and disk8 dropping? Also, any writes to the array in the meantime?
May 31, 20251 yr Author I'm not totally sure, I hadn't got around to fixing my email notifications. Everything was good 2-3 weeks ago when I was in there. I was extremely low on free space but that's all I noticed. I'm assuming that drive 5 dropped out first because it is showing as emulated but I suppose its possible that they were both affected by the same power failure. If they happened at different times, It's possible some things were written to the array in between but I had those older drives excluded from the most used shares so it's also possible that there was not.
June 1, 20251 yr Community Expert Disk5 appears to be failing, you can try to recover the most data possible from it by cloning it with ddrescue, and then use that to try to rebuild disk8, but success will depend mostly on how much it can be recovered from dis5, and if there weren't any changes to the array after it got disabled.
June 1, 20251 yr Author Makes sense. Drive 5 mounts fine on a linux VM and the data looks intact for now. Is the only way to force #5 back into the array to rebuild #8 with it is to do the ddrescue clone to a new disk and then update the config with the new drive?
June 2, 20251 yr Community Expert 15 hours ago, Jparlin said:Is the only way to force #5 back into the array to rebuild #8 with it is to do the ddrescue clone to a new disk and then update the config with the new drive?That's what I would recommend, if enable current disk5, there will likely be more errors duding the rebuild
June 2, 20251 yr Author Ok. I have cloned drive 5. No read errors at all. When adding the new cloned drive in it's place, the array shows it as new though and wont start the array to allow a rebuild. Is there a way for me to set it as the existing drive and not new?
June 2, 20251 yr Community Expert Do this:-Tools -> New Config -> Retain current configuration: All -> Apply-Check all assignments and assign any missing disk(s) if needed, including the cloned disk5 and a new disk8, replacement disk should be same size or larger than the old one-IMPORTANT - Check both "parity is already valid" and "maintenance mode" and start the array (note that the GUI will still show that data on parity disk(s) will be overwritten, this is normal as it doesn't account for the checkbox, but it won't be as long as it's checked)-Stop array-Unassign disk8 -Start array (in normal mode now), and post new diagnostics
June 3, 20251 yr Author Everything went smoothly from the GUI. Here is the new Diagnostic. tower-diagnostics-20250602-2057.zip
June 3, 20251 yr Community Expert Check filesystem on the emulated disk8Unassign sdn is spamming the log with errors, which disk was this?
June 3, 20251 yr Author Perfect! I had to use xfs_repair -L to trash the log file but it came back quickly. Most file and folders are still intact. There are some in lost and found but the sub folder and file names look intact. Not sure if I will keep those or restore them from backup instead.sdn was a potential new drive that failed preclear. It's getting RMAd, I pulled it out. tower-diagnostics-20250603-0915.zip
June 3, 20251 yr Community Expert Looks good, if you are happy with disk8 contents, you can just rebuild to a new disk.
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.