treesap

Thanks for your help. After you suggested that the data was actually missing entirely, I realized I could compare with a Dynamix File Integrity export. Unfortunately, this confirmed the disk of data was definitely missing. I appreciate you sharing the link. The documentation has been significantly updated from my previous looks and the pages I found while searching! It's looking like I did the equivalent of the filesystem check via manual commands, but now I know how to do so while preserving parity (as well as a GUI way!) I'll be sure to reference it in the future. Thanks again.

Hi there, I've used Unraid for several years and have had to recover from various situations and have generally succeeded, but I need some help with my current one. I had a disk (disk1) fail and show unmountable (I think due to connector issues). I replaced cable and let it rebuild, but the rebuilt drive still showed "Unmountable". I removed it from the array and ran an xfs_repair on it, but it failed to find any secondary superblocks and the process failed. I've tried doing a manual preclear, then add it. The disk is always mountable until I add it to the array, then rebuild makes it unmountable. BTW, at one point I tried a different spare drive, but same result. I thought an easy fix would be to just move the important files to other disks manually, then reformat the disk, but there's no /mnt/disk1 directory to move from. I can't tell if this is expected or not. I think the emulation is working correctly for shares themselves—or at least I haven't noticed anything missing—but I could be wrong. I'll note that when messing with this, I accidentally unplugged a 2nd disk while my disk was in this "unmountable" state, and it was able to utilize this unmountable one to successfully restore it. Anyway, advice for fixing this is much appreciated. It seems like just moving files off of it would be a good option, if it's reasonable to do so. An idea came to mind for a script that scans the shares for files that don't have the device IDs of the "working" drives or that aren't present in the /mnt/disk folder, but it feels precarious and would take some experimentation, so I thought I'd check with the experts first. Hopefully there's a nicer way. Attaching diagnostics. Currently the disk is rebuilding (I wanted insurance for other drive failures, since that apparently still functions. And yes, some of the drives have a ton of hours—planning to replace with fewer larger drives before long.) Thanks in advance, so much! tower-diagnostics-20241202-0930.zip

I'm really hoping somebody has something to share. I've tried all the things suggested in this thread and the seemingly related issues I've read through on GitHub. BTW, I'm pretty sure the weird characters just represent progress of RSA key generation, so the issue seems to be that we're getting stuck somewhere just after that. The Calibre container is the one that I only update when I'm fully prepared to spend hours troubleshooting. I should really just some extra hours to make my own personal Calibre container that I know everything about (and can thus fix)!