I have an UNRAID setup with (1) 12TB parity, (2) 12TB disks and (1) 2TB disk.

The system has been on battery backup and running for almost 500 days. I needed more storage, so I ordered another 12TB drive. I used the pre clear plugin and after almost 2 full days, it was at 98%. When I went to check on it later that night, the UNRAID GUI was acting weird, nothing was loading, I couldn't stop the array, I wasn't able to SSH into the server (it's remote).

I left it overnight and found it the same in the morning. I had someone physically power it down. I started the array and noticed the the 2TB drive was showing an unhealthy SMART icon. I was already planning on moving this drive out once I got the new 12TB drive up. The parity check was around 8% when I went to bed, when I woke up, I wasn't able to remote in. The UNRAID was unresponsive.

I got the server back up and left it stopped. I ran a short smart test on the 2TB and threw the logs into chatGTP. ChatGPT said that based off of the logs, the drive is failing, has been for awhile and would explain the freezing during the pre clear and parity check.. it suggested to remove the 2TB drive, assign the new 12TB in it's place, then rebuild the 2TB to the 12TB, then after that, run the parity check.

I've had ChatGPT burn me before, so I'm checking in on the experts. Is this good advise?

I wasn't able to find any previous logs because I guess these are stored in RAM? I can setup a syslog server if anyone highly recommends me to? I know these freezes are typically hardware related.. I'm just looking for the best advise. Thanks

Please post the diagnostics to check SMART, and note that a bad drive doesn't typically crash a server, so there may be something else going on, even if the drive is bad.

btch-smart-20260106-1102.zip

Please read the Diagnostics link to understand what we are asking for.

Attach Diagnostics ZIP to your NEXT post in this thread.

Ohh sorry!

btch-diagnostics-20260106-1445.zip

1 hour ago, Seanmc980 said:

I was already planning on moving this drive out once I got the new 12TB drive up.

Why not just replace the 2TB disk2 with the new 12TB and rebuild?

Can't tell anything about your filesystems since those diagnostics were taken without the array started.

Start the array and...

Attach Diagnostics ZIP to your NEXT post in this thread.

1 minute ago, trurl said:

Why not just replace the 2TB disk2 with the new 12TB and rebuild?

Can't tell anything about your filesystems since those diagnostics were taken without the array started.

Start the array and...

Attach Diagnostics ZIP to your NEXT post in this thread.

My concern is that the server crashed during the pre clear of the new drive, then crashed during the parity check.. I didn't want to get my server into a position where I was trying to do too much at once and make matters worse.

I did shut the server down, reseated all connections, removed several unassigned drives that were there just to move files around.

I'd like to just swap the failing 2TB with the new 12TB, but wanted to be sure this was the right approach.. I didn't want to start the array yet, because the parity check had previously failed. I'm just looking for the best advise to move forward.

Set the 2TB disk2 to not assigned, then start the array. That will allow us to see if parity can emulate the missing disk. Then...

Attach Diagnostics ZIP to your NEXT post in this thread.

Ok, I clicked on the failing disk (2) and removed it. Started the array (had to check the box) and ran the diagnostics btch-diagnostics-20260106-1649.zip

Emulated disk2 shows plenty of data. You can browse its files if you want (you can even write to it).

Should be OK to rebuild to the new 12TB. Assign new 12TB as disk2 and start the array to begin rebuild.

Thanks Trurl! Giving it a go now.

The disk replacement went smooth. It was rebuilt in about 12 hours. I assumed, after the rebuild, that it would run a parity check, considering the freeze during the initial pre clear and the freeze during the following parity check.. the array shows started and it's states that parity is valid. It shows that the last parity check was today at 8:01am.

This is confusing because the parity checks alone have typically taken my system over 24 hours.. I'm attaching my latest diagnostics. Should I just run a parity check, or am I good??

Thanks!btch-diagnostics-20260107-1203.zip

10 minutes ago, Seanmc980 said:

The disk replacement went smooth. It was rebuilt in about 12 hours. I assumed, after the rebuild, that it would run a parity check

The rebuild process uses the combination of parity plus other drives to work out what to write to the drive being rebuilt. This means that if there were no errors during the rebuild then parity is automatically in sync.