I’m running Unraid 7.1.2 with 5 data drives and 1 parity drive.

A few days ago, during a parity sync, my server unexpectedly restarted.
When it came back online, Disk 2 showed a warning: "Disabled – contents emulated."

Following advice from the forums, I:

1. Removed the disk from the array.

2. Started the array without it.

3. Stopped the array again.

4. Re-added the drive to the array.

When I started the array, the disabled warning was gone, and the system began rebuilding the drive.

Initial rebuild behavior:

• Speed: ~200–300 MB/s

• Progress: smooth until ~98%

• At 98%: speed dropped to ~1–2 MB/s

• At 99%: dropped to just a few B/s, with an ETA of >1000 days.

Someone on the forums suggested the slowdown might be due to the drive being written to during rebuild. They recommended running the rebuild in Maintenance Mode.

What I tried next:

• Stopped the array.

• Started it again in Maintenance Mode.

• Restarted the rebuild.

Current problem:
Now the rebuild is only running at 1–3 MB/s from the very start.

Additional notes:

• Drives are ~2 years old.

• Low usage overall.

• No obvious SMART errors so far (but I can post a report if needed).

Has anyone experienced something similar, or have ideas on how to restore normal rebuild speed? Could this be a hardware issue, or something with Unraid after the restart?

Thanks in advance!

linux-node-00-diagnostics-20250813-0737.zip

Edited August 13, 2025Aug 13 by GooseDeer
better formatting, and clearer message

Seeing a lot of errors related to the controller, but since it's a RAID controller unclear if that's the problem or an issue with a device:

Aug 13 00:11:35 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33889 (808355494s/0x0002/FATAL) - Unrecoverable medium error during recovery on PD 04(e0x20/s4) at 565ea5d40

Aug 13 00:11:35 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33890 (808355494s/0x0001/FATAL) - VD bad block table on VD 04/2 is full; unable to log block 565ea5d40 (on PD 04(e0x20/s4) at 565ea5d40)

Aug 13 00:11:40 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33892 (808355500s/0x0002/FATAL) - Unrecoverable medium error during recovery on PD 04(e0x20/s4) at 565ea5e48

Aug 13 00:11:41 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33893 (808355500s/0x0001/FATAL) - VD bad block table on VD 04/2 is full; unable to log block 565ea5e48 (on PD 04(e0x20/s4) at 565ea5e48)

Aug 13 00:11:44 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33895 (808355504s/0x0002/FATAL) - Unrecoverable medium error during recovery on PD 04(e0x20/s4) at 565ea5da8

Aug 13 00:11:44 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33896 (808355504s/0x0001/FATAL) - VD bad block table on VD 04/2 is full; unable to log block 565ea5da8 (on PD 04(e0x20/s4) at 565ea5da8)

Aug 13 00:30:12 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33914 (808356612s/0x0002/FATAL) - Unrecoverable medium error during recovery on PD 04(e0x20/s4) at 5662195c8

Aug 13 00:30:12 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33915 (808356612s/0x0001/FATAL) - VD bad block table on VD 04/2 is full; unable to log block 5662195c8 (on PD 04(e0x20/s4) at 5662195c8)

Aug 13 00:30:57 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33923 (808356656s/0x0002/FATAL) - Unrecoverable medium error during recovery on PD 04(e0x20/s4) at 56621dc70

Aug 13 00:30:57 Linux-Node-00 kernel: megaraid_sas 0000:08:00.0: 33924 (808356656s/0x0001/FATAL) - VD bad block table on VD 04/2 is full; unable to log block 56621dc70 (on PD 04(e0x20/s4) at 56621dc70)

Try again and post new diags if it happens again, but note that RAID controllers are not recommended with Unraid

I didn't know that raid controller would cause issues. I currently use a Dell T320 as my server and it has a PERC H710 controller. I have just purchased a H310 in IT mode to replace this with.

In the meantime, I will run the array in maintenance mode, and start the data-rebuild again and provide new diag files if it grinds to a holt again.

Hello, So I bought a H310 to replace my H710. I have switched everything over and I have started the parity-check again.

Switching over to the H310 provides me with a lot more information and not understand why the raid controller was not recommended. As stated, before I have 6 disks with are 12 TB with one being the parity. 2 of the drives were bought new, and the other 4 were bought refurbished a year ago. I didn't test these drives when I bought them, but they seem to have been working up until now. However, I cannot say that for sure as a lot of the information I am seeing was hidden being the H710 raid controller. I can see the following drives have errors:

Disk 1:
Reallocated sector count: 27720
Reported uncorrect: 2743

Disk 3:

Reallocated sector count: 7200
Reported uncorrect: 2380

The status of the party-check is:

Total size:	12 TB
Elapsed time:	22 hours, 25 minutes
Current position:	11.8 TB (98.2 %)
Estimated speed:	987.0 KB/sec
Estimated finish:	2 days, 12 hours, 33 minutes
Sync errors corrected:	60

Once it gets to around 97% is drops to like 1-2MB/sec, and I have seen it drop to low B/sec when it gets to ~99.7%

I have attached a new set of diagnostic logs.

I have a couple questions:

1. With the errors, does this mean that the party check is pointless as it ends up taking too long?

2. Are these large error numbers for Disk 1 and Disk 3? I don't really have a reference point.

3. Should I replace these drives?

4. Would it be possible to get a refund from amazon for these drives as I bought them a year ago this month?

5. If there any next steps I can do to improve my situation without have to buy new drives?

linux-node-00-diagnostics-20250818-1953.zip

Disk3 appears to be failing; run an extended SMART test, it may also be worth it to run one on any other disk with suspect SMART, like disk1