HDD failed during parity check

[Nexius2]

Hello, a parity check started Monday morning and it detected a HDD failure. no problem, I have 2 parity drives and I ordered a new HDD for replacement.

but I have noticed the parity check stays at 19.8% since last night. since I'll have to stop the array to replace the HDD is it something normal? will it end eventually or no chances?

thanks

nostromo-diagnostics-20221220-1500.zip

[Vr2Io]

31 minutes ago, Nexius2 said:

will it end eventually

Won't,  because disk failure and no reason to waiting it end.

Today, I also replace a pre-fail data disk and a trouble SSD ( only use 3 month+ ).

Edited December 20, 2022 by Vr2Io

[trurl]

Looks like disk19 (56NT) disconnected as sdc and reconnected as sdac. SMART for 56NT looks fine, emulated disk19 is mounted and has plenty of contents.

 

Your parity check paused after that at 18:00, resumed at 00:00. Are you sure it isn't just paused now? Post a screenshot of Main - Array Operation.

 

Should be fine to rebuild to the same disk after checking connections. Slightly safer to rebuild to a new disk since you can keep the original with its contents intact in case of problems.

 

 

[trurl]

2 minutes ago, Vr2Io said:

Won't,  because disk failure and no reason to waiting it end.

With dual parity and only a single disk disabled it should be able to complete parity check.

 

Looks like you added new disk20 and it was clearing before parity check started. Some other things in syslog related to this but I haven't looked at it all closely yet. Can you tell us more about this and how it relates?

 

Not sure how Unraid would handle a new disk and a disabled disk with dual parity, maybe you can't complete parity check like that.

[trurl]

Are you trying to add disks under power? Hotswapping not recommended with array or pool disks even if hardware supports it. Unraid won't do anything with a new disk until you assign it, and you can't change assignments without stopping the array, so not much point in hotswapping.

 

Possibly you disturbed connection to disk19 when installing disk20.

[Vr2Io]

30 minutes ago, trurl said:

Post a screenshot of Main - Array Operation.

That's fine, check does "estimated finish" decreasing time to time.

 

[Nexius2]

so, I added disk20 a couple days ago. no hot plug so I stoppd the server to add it. array was fine nothing to notice after array growth.

and 1 or 2 days after the parity check started (it stops at 18h and starts back at midnight).

during the check it seems to have found errors on disk 19. since then, parity check has stopped, but I can see data move between all the HDD as if it was still going on.

I just notice sdac has appeared (thanks to you) don't know if it's good, it will be changed anyways.

maybe some power plug has moved while adding disk20....

 

but nothing realy explains why parity is stuck at 19.8%

 

 

[Nexius2]

[trurl]

8 minutes ago, Vr2Io said:

That's fine, check does "estimated finish" decreasing time to time.

Sorry, I meant screenshot from OP. If you have a problem you should start your own thread.

[trurl]

Just now, Nexius2 said:

Try again without clipping some of the parts I wanted to see.

[trurl]

3 minutes ago, Nexius2 said:

parity check has stopped, but I can see data move between all the HDD as if it was still going on.

Don't know what you mean by that. Parity check doesn't move any data. It reads all disks, and if it is a correcting check (I don't think yours was) it will write corrections to parity. With a non-correcting parity check, any writes must be coming from another process.

[Nexius2]

is this what you want?

 

[trurl]

No, I want only Main - Array Operations, but all of Main - Array Operations. You clipped the top and bottom of that section when you posted it before.

[Vr2Io]

5 minutes ago, trurl said:

Sorry, I meant screenshot from OP.

I am not provide the screen for you, OP should check that to decide next step.

Sorry to interrupt "your" thread.

[Nexius2]

[trurl]

Also, FYI, your post of Main - Array Devices is also clipped on the right, making it impossible for me to see that emulated disk19 is mounted and has contents.

[trurl]

2 minutes ago, Vr2Io said:

I am not provide the screen for you, OP should check that to decide next step.

Sorry to interrupt "your" thread.

I figured you joined the thread to help. Just got confused about your screenshot.

[Nexius2]

every time I had this, it was written emulated but not this time....

but it seems to be.

[trurl]

1 minute ago, trurl said:

Also, FYI, your post of Main - Array Devices is also clipped on the right, making it impossible for me to see that emulated disk19 is mounted and has contents.

I don't need to see it though since I already could determine that from diagnostics.

 

[Vr2Io]

Fine.

 

I just want to confirm does corrected parity check still in progress. If nothing have change it likely in stop. ( I currently in rebuild, finish time will change continue )

 

   

 

 

Edited December 20, 2022 by Vr2Io

[Nexius2]

Just now, Vr2Io said:

Fine.

 

I just want to confirm does corrected parity check still in progress. If nothing have change it likely in stop.

 

 

 

nothing has changed since last night

[Vr2Io]

Then you should restart rebuild by new disk or same disk. And must confirm the power connector in good connect state.

[trurl]

6 minutes ago, Nexius2 said:

every time I had this, it was written emulated but not this time....

but it seems to be.

That does show emulated disk19 is mounted and has plenty of contents. Not sure what you mean by "written emulated". If you mouseover the red X it will tell you the drive is emulated. And emulated disks can be written to (by updating parity) so those emulated writes can be recovered. If some other process was writing to disk19 those writes could be recovered by rebuild.

 

Normally, parity check won't write any data disk. But Unraid only disables a disk when a write to it fails.

 

If a read fails, Unraid will try to get the emulated data from the parity calculation and try to write it back to the disk. If that write fails the disk is disabled. So parity check can disable a data disk even if nothing else is writing to it.

[Nexius2]

thanks, I'll let it do it's thing until I receive the new HDD and will stop it then to replace HDD and check connections.

[trurl]

28 minutes ago, Nexius2 said:

I added disk20 a couple days ago

Did you format it? It doesn't appear to have mounted in diagnostics so I assumed it hadn't been formatted yet.

 

29 minutes ago, Nexius2 said:

just notice sdac has appeared (thanks to you) don't know if it's good, it will be changed anyways.

Yes, that Unassigned Device is the disk formerly assigned as disk19. SMART for it looked OK to me so you could rebuild to it if you wanted to. Do you know how to rebuild to the same disk?

 

If you want to wait on the replacement to rebuild

1 hour ago, trurl said:

Slightly safer to rebuild to a new disk since you can keep the original with its contents intact in case of problems.

don't do anything to that original disk until you are satisfied with the results.