I'm a bit in a squeeze, I have a array that consists of the following:

 

P - WD12TB

D - WD12TB - disabled because of read/write errors.

D - MX500

D - MX500

 

I have ordered a 18TB disk to replace the disk with the IO-errors but found out that I have to replace the parity drive because the replacement is bigger than the data-drive. Can I do the following and keep all of my data?

 

1. Stop the array

2. Power down the system

3. Replace the faulty drive with the 18TB

4. Switch the old parity drive to the data-slot

5. assign the 18TB drive to the parity-slot

6, start the array

You can use this:

https://wiki.unraid.net/The_parity_swap_procedure

7 minutes ago, FDMK said:

D - MX500

D - MX500

SSDs in the array cannot be trimmed, and can only be written at parity speed.

 

8 minutes ago, FDMK said:

D - WD12TB - disabled because of read/write errors.

Might be a good idea to post diagnostics so we can see why and also see if the emulated disk is mountable.

5 minutes ago, JorgeB said:

You can use this:

https://wiki.unraid.net/The_parity_swap_procedure

Yeah, I have read that article but I wasnt sure it was applicable to my situation.
 

@trurl I can consume the data that is on the share that is hosted on the 12TBs.

5 minutes ago, FDMK said:

I can consume the data that is on the share that is hosted on the 12TBs.

Not entirely clear what "consume" means in this context but no matter.

 

Do you intend to rebuild the disk or not? Is the emulated disk mountable?

1 minute ago, trurl said:

Not entirely clear what "consume" means in this context but no matter.

 

Do you intend to rebuild the disk or not? Is the emulated disk mountable?

I intend to keep my data The way I see it is that the parity-disk contains all of my data, no? Can I execute the plan I wrote?
What I meant by 'consume' is that I can play the media that is on the share. I will post the diag.

3 minutes ago, FDMK said:

The way I see it is that the parity-disk contains all of my data, no?

NO

 

The parity disk contains NONE of your data.

 

https://wiki.unraid.net/Manual/Overview#Parity-Protected_Array

 

 

5 minutes ago, FDMK said:

Can I execute the plan I wrote?

You should execute the parity swap procedure.

6 minutes ago, FDMK said:

can play the media that is on the share

If all the files for that share are on the emulated disk then it is mountable, but if any are on other disks who knows?

Diag attached

fdmnas-diagnostics-20230427-1047.zip

Emulated disk1 is mounted. Physical disk1 is disconnected. No evidence there is anything wrong with the disk, because there is no evidence of the disk itself. Syslog seems to indicate it was probably just a connection problem though.

 

Check connections and post new diagnostics

I replaced the cabling, performed an extended smart test and a FS-check/repair, everything seems to be working fine according to the logs. I decided to rebuild the original data-drive before replacing the parity disk with the 18TB.

Maybe not, after a day the disk errors again 😞

 

Apr 30 11:18:58  kernel: md: disk1 read error, sector=13464173520
Apr 30 11:19:08  kernel: md: disk1 write error, sector=13464173520

post new diagnostics

fdmnas-diagnostics-20230501-0638.zip

Apr 29 23:44:10 FDMNAS kernel: ata3: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 29 23:44:10 FDMNAS kernel: ata3.00: revalidation failed (errno=-2)
Apr 29 23:44:16 FDMNAS kernel: ata3: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 29 23:44:16 FDMNAS kernel: ata3.00: revalidation failed (errno=-2)
Apr 29 23:44:22 FDMNAS kernel: ata3: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 29 23:44:22 FDMNAS kernel: ata3.00: revalidation failed (errno=-2)
Apr 29 23:44:22 FDMNAS kernel: ata3.00: disable device
Apr 29 23:44:28 FDMNAS kernel: ata3: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 29 23:44:28 FDMNAS kernel: ata3.00: detaching (SCSI 3:0:0:0)

 

Disk dropped offline, did you replace both power and SATA cables? If yes swap ports with a different disk, if it happens again to the same disk it's likely a disk problem.

It has been a while. Took a lot of testing and rebuilding but I think I have pinpointed the issue now.

I applied the tunables from powertop and that seems to be the issue, especially the 'Enable SATA link power management', when those are applied it takes a couple of days and/or disk sleep cycles to 'fail' the disk.