Disk Error, Last Parity Canceled - How to replace disk?

[lanky8804]

So I have the current scenario and before I go messing with anything was hoping for some advice please.

 

Last parity check got stuck at 3.4% after 16hours due to a disk with errors (can hear it clicking away). Disk still green and no failure notice as yet.

 

As a result I had to cancel the latest parity check (28/3/21) however I run parity checks weekly and the one before this was successful (21/3/21)

 

What is the best way to replace this disk?

If you cancel a parity check does it revert to the last fully complete one?

 

Many thanks

[trurl]

5 minutes ago, lanky8804 said:

I run parity checks weekly

Why? Parity checks are only needed occasionally as a test to verify that parity is still valid. Unraid parity is realtime, every write to the array updates parity at the same time so parity should always be valid. Most people schedule them monthly but some less frequently.

 

And so, this is meaningless:

5 minutes ago, lanky8804 said:

does it revert to the last fully complete one?

 

  

5 minutes ago, lanky8804 said:

hoping for some advice please.

 

If possible before rebooting and preferably with the array started
Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.

[lanky8804]

@trurl Ah ok, so I presume with this information monthly if not greater would be acceptable for parity then moving forward.

 

Full Diagnostic attached.

unraid-diagnostics-20210329-1858.zip

[trurl]

Do you actually have any SMART warnings on the Dashboard page for any disk?

 

Looks like a connection issue with disk4.

 

Also you have a RAID controller. Are you actually using it? Not recommended.

 

Why are your disks still using ReiserFS?

[lanky8804]

@trurl

No, no SMART warnings just a bunch of errors on read/write that and it make a clicking noise whereas it never use to. 
 

odd about connection issue as haven’t opened in months now. 
 

yes have a RAID card however has been flashed with firmware so just runs as JBOD with no RAID. This is all left to UnRAID as recommended. 
 

still in old format as wasn’t sure how to upgrade then without loosing data and I don’t have the space elsewhere to move data then re-format, any ideas on this would be good also please if I have missed something simple. 
 

so should I be all Ok just to replace a disk as normal, unmount pool, deselect disk, shutdown, replace etc?

 

[trurl]

Not clear the disk needs replacing. 

 

Run an extended SMART test on it 

[lanky8804]

ok will run one and post once complete

[lanky8804]

@trurl Have tried a few times to run an extended SMART test but will not complete. Attached is the latest log for Disk 4.

ST3000DM001-1CH166_Z1F3746X-20210329-2027.txt

[trurl]

# 1  Extended offline    Interrupted (host reset)      00%     43310         -

That indicates you did not let it complete. How long did you let it run? I usually estimate 2-3 hours per TB for any operation that needs to work on the complete disk.

[JorgeB]

If the drive is assigned to the array make sure to disable spin down during the test.

[lanky8804]

Thank you will try again, I left it running overnight. Maybe it tried to spin down.

[lanky8804]

Nope just tried again and failed after 1 hour saying 'interupted (host reset)' Obviously I didnt cancel it and disk was now set to not spin down.

[trurl]

Well, you wanted to replace because it was clicking and you are probably right. And it is pretty old.

On 3/29/2021 at 2:55 PM, lanky8804 said:

replace a disk as normal, unmount pool, deselect disk, shutdown, replace etc?

All you need to do is assign a new disk to that same slot and start the array to begin rebuild. No need to deselect since it will know a new disk is there when you assign it. Deselect is only needed when you want to reassign the same disk and make it rebuild.