Hi All

I got home tonight to find my server was showing read errors on 3 disks!!!

This is what the dash showed...

All my dockers were disabled. So, I downloaded the diags (attached) and then knowing that a reboot usually fixes most problems I rebooted it.

It now shows that the parity disk is disabled... so, I am guessing it will need to rebuild (or check) the parity disk from the data drives... but, I am not sure what to do from here and am very twitchy about not having a working parity disk enabled.

Can anyone advise what to do from here please?

Thanks

hector-2-diagnostics-20260621-1901.zip

Would I be correct in thinking that I should;

1. Stop the array

2. Unassign the existing parity disk

3. Start the array to allow UnRAID to see the change

4. Stop the array

5. Assign the originally disabled disk as the parity drive

6. Start the array and allow a parity build on the 'new' disk

I just dont want to start something which is going to cause me an issue going forward

Thanks

Jun 20 13:11:02 Hector-2 kernel: hpsa 0000:0a:00.0: Controller lockup detected: 0xffff0000 after 30

Jun 20 13:11:02 Hector-2 kernel: hpsa 0000:0a:00.0: Telling controller to do a CHKPT

Jun 20 13:11:02 Hector-2 kernel: hpsa 0000:0a:00.0: controller lockup detected: LUN:0000000000800001 CDB:01030000000000000000000000000000

Jun 20 13:11:02 Hector-2 kernel: hpsa 0000:0a:00.0: Controller lockup detected during reset wait

Jun 20 13:11:02 Hector-2 kernel: hpsa 0000:0a:00.0: failed 28 commands in fail_all

Problem with the controller. Make sure it's well seated and suficially cooled, you can also try a different PCIe slot if available.

1 minute ago, SliMat said:

1. Stop the array

2. Unassign the existing parity disk

3. Start the array to allow UnRAID to see the change

4. Stop the array

5. Assign the originally disabled disk as the parity drive

6. Start the array and allow a parity build on the 'new' disk

You can do that

Thanks @JorgeB

As it was showing several disks as failed, I went into the room where the server is and it was quite warm, so I turned on the aircon and have left it on, so suspect that it is a heat issue.

Assuming that is the cause would my de-assign then re-assign routine be correct?

Thanks

Yes, already mentioned it above.

Just now, JorgeB said:

Yes, already mentioned it above.

Thanks, I read it as "you could do that" - and thought there might be a better way ;)

The server I am using is an HP ProLiant ML30 Gen10 - I think I fitted an HP P410 RAID card and switched it to JBOD mode... is there a better (newer) SATA controller which anyone can recommend?

Thanks again

You can see here: https://forums.unraid.net/topic/102010-recommended-controllers-for-unraid/