I believe I'm about to experience my first data loss with UnRaid since I bought my key in 2006. And it's all my own fault 

 

My Disk 2 was being rebuilt (my fault, detailed in another thread) and during the rebuild, the server went unresponsive (to login's either local or via SSH). I also think I know why this might have happened, detailed in my "Post Mortem" below the image. So, likely, my fault.

 

I hoped it was still rebuilding and that a service went down or something... but 15 hours later (long after a rebuild would have been complete), I decided to reboot. Disk 2 did not rebuild, and my two old IDE drives (which I was planning on replacing this weekend and why I'm working in disks') are now saying "Wrong". But they are the right disks.

 

I think I'm dead in the water. Should I try to read the data off the two 400 GB drives? In essence I'd just be losing the data from the 1TB drive... Then, if I do that, do I change all 3 of these drives to "no device"? But somehow I'd have to force UnRaid to rebuild the parity without those 3 devices...

 

Post Mortem, I think I know why the two IDE drives might have gone offline during the rebuild. I had to plug in 1 fan while the server was rebuilding, and might have jostled that $*&!@ finicky IDE cable so that the drives went offline. That sounds about like my luck so far.

 

Anyway, thoughts would be welcome. My gut reaction is to take the 400GB IDE drives to my WinPC and try to read the data off of them.

Yeah, pretty sure the IDE drives were disconnected during the rebuild. "Partition format: Unknown". Probably nothing I can do short of data recovery. Maybe an old version of TestDisk that supports ReiserFS...

 

 

If it's the RFS superblock and if reiserfsck --rebuild-sb[X] worked, is it possible UnRaid could recognize the disks again?

 

I'm tempted, but don't want to run anything on those disks until someone who knows what their doing says it's worth a try...

Did you grab diagnostics before you rebooted? I wouldn't move disks to your Windows PC as natively it won't be able to read them, and isn't that how you managed to break Disk 2 in the first place? I can't help with ReiserFS, I'm afraid, as I haven't really used it seriously. Of all the file systems though, it seems to be one that people have had a lot of success in recovering from even very severe corruption.

If, as you say, your IDE drives became disconnected during the rebuild of Disk 2, they would be being read at the time so I wouldn't expect the corruption to be severe. Keep them in the array and see what others suggest and you still might be able to recover.

Okay, thanks for your insight, John. That sounds like a reasonable idea. Regarding logs, unfortunately, I was forced to power cycle reboot once it went unresponsive.

 

You're right regarding being a bad idea moving them to a WinPC. There are some good tools in Linux, so I'll leave the drives in the UnRaid server.

 

Just doing some "checks" I got the dreaded "Unknown code er3k 127" message. I'll continue searching the forum and see if any more responses come in.

 

Ugh... by the look of some of these logs, I might have fried those old IDE drives.

tower-diagnostics-20181123-1023.zip

Try replacing the IDE cable.

Interesting idea! Trying now.

OMG! YES!!!
THANK YOU, THANK YOU!!!

Tears of happiness. I NEVER would have though about an IDE cable going bad.

 

Drive 2 is once again, happily being rebuilt and that IDE cable is in the trash.

 

Let me know how I can buy you a beer.

Glad it worked, just make sure you get rid of those IDE drives, problem with one can affect the other, because of the master/slave issue with IDE mode.

IDE drives are both officially replaced! Yeah, I left them both attached as-is, to maintain proper master/slave as each one was swapped out. Only after both were swapped did I pull both of them.

Thanks again.