I'm running unRAID OS 5.0-rc5

 

I had a disk that got disabled. I was running 3 1.5 TB Seagate drives, 1 parity and 2 data. I bought a new WD 2TB drive and needed to upgrade the parity drive to 2TB and use the old parity as a replacement for the bad data drive.

 

I powered everything down, swapped the wrong drive to the new one so I powered down again and this time I hooked them up as I understand they are supposed to be. That is: Old parity drive sata cables hooked to new 2TB drive and bad data drive sata cables to old parity drive. Leaving the disabled bad drive unhooked.

 

Started back up and now the parity drive is (sda) where before it was (sdb) and my first data drive (the one that was bad) shows up as Not Installed. I decided that something is most certainly not right with that and I would shut it down again, put everything back exactly as it was, ensure I shut down the array properly, turn off several of my plugins and check everything looks good and then retry swapping the drives as I did before. Upon booting back up everything still looks as it did with the 2 TB drive hooked up (but it's not), Parity is now sda still and disk 1 is Not Installed. Also noted quite a few reads and writes (12,783 and 22,398) on parity so now I've freaked out a bit and I'm trying to stop the array and it's just showing "Retry unmounting user share(s)..." Hopefully I haven't lost everything.

 

Is there a chance that the parity drive switched to the disabled data drive and started rebuilding itself? Hopefully not. I'm thinking at this point since I don't have a disabled drive but one that's "Not Installed" I just need to go through the process of upgrading the parity drive to 2 TB drive then set the old parity drive up as a data disk and hook my bad drive up to an external and copy as much as I can to the old parity now new data drive?

Since I had everything back up and was showing the old drive as not installed I went and made sure all my plugins were disabled and would not run at a restart. At this point I had to kill a few processes, namely for the itunes server since it didn't stop and it would not allow me to Stop the Array. I just kept seeing Retry unmounting user shares. If anyone sees this it's handy to do a ps aux | more and look for any processes that unraid didn't start and kill them. After the array was stopped I could assign the bad drive into the slot again and I'm rebuilding data now. I figure the worse that could happen is unraid disables it again and I can try the swap disabled procedure again following the steps more closely. My guess is I didn't stop the array when I should have somewhere when I was rebooting. I still welcome any other guidance

You should have hooked the new drive up and used the pull-down tabs to assign the drives how you wanted. Just connecting them to the SATA cables doesn't mean much.

 

At this point I think I might consider short stroking the 2TB down to 1.5 using the same procedure that is used to remove HPA, only setting the size to be identical to your existing 1.5 drives. That way you can rebuild the failed drive onto the new drive without expanding parity. You wouldn't be gaining any space right now anyway, since only the parity drive would be a 2TB. The next drive purchase could then be done at your leisure, and be a 2TB or larger, and assign it as parity. After the array is fully healthy with a larger parity drive, then you could resize the current 2TB back to full size and rebuild onto itself to gain the 500GB.

Thank you lionelhutz and jonathanm. jonathanm: that certainly makes a lot of sense, I'll see how this rebuild goes and how the drive does over the next week or so. lionelhutz: perhaps that was the mistake I was making. Like I was saying, maybe I'll get the chance to try swap disabled again.