New around here, but hoping someone have the time to help an ill-prepared soul out.

 

Here's my setup (before the datadrive failed):

[parity] 1x 4TB Seagate IronWolf 5900rpm

[data] 2x 4TB Seagate IronWolf 5900rpm

[data] 2x 3TB Seagate ironWolf 5900rpm

[data] 2x 2TB Seagate Barracuda 7200rpm

[cache] 1x 240GB Corsair SSD

 

SSD connected to the motherboard, while the harddrives are connected to an Areca controllercard.

 

Everything running on a Ryzen-based system.

 

The plan was to swap out the parity for a 8TB IronWolf 7200rpm, and, after rebuilding the parity, swap out one of the 2TB-drives for another 8TB IronWolf.

Unfortunately, while rebuilding parity on the new drive, one of the 3TB drives failed...

While testing, it passed the SMART-test, but failed the Short-Generic test (and now does not even show up anymore).

Tested on a separate computer with linux.

 

My problem now is as follows:

I have two brand new 8TB disks, but my old parity disk was only 4TB. Is there a way I could get the array up and running again with what I have on hand?

Waiting a couple of weeks for the warranty replacement to arrive is something I'd rather avoid right now, if at all possible. 😕

 

Thanks in advance.

 

edit: It seems to me that this should be doable using the parity swap procedure. Can someone confirm/hold my hand a bit here? 🙏😊

 

Testlog_ironwolf.txt

Edited December 2, 20187 yr by DogeKitteh007
Learned to read

Go to Tools-diagnostics and attach the complete zip to your next post. 

39 minutes ago, trurl said:

Go to Tools-diagnostics and attach the complete zip to your next post. 

Attached the diagnostics. Strangely enough, disk 3 now returned to normal operation.

I stopped the parity-build. Estimated time was now around 5 days(!). Last time it started, it read a lot faster (est. 2 days).

 

Thinking about using the parity swap-procedure.

At the moment, it's a 8TB drive (parity not built) selected as parity-drive.

 

I seem to be unable to start the array when following the procedure though.

"Too many wrong and/or missing disks!" after unassigning the datadrive (3TB disk3)

Could this be a change since 6.1?

Not sure if I should just shutdown and give the swap a try anyway.

 

unraid-diagnostics-20181202-0208.zip

Edited December 2, 20187 yr by DogeKitteh007
Parity swap procedure not updated for 6.6.5?

You shouldn't do anything without further advice.

I've shutdown the server for now. 

It's 4am here. Getting tired.

Think I need some assistance anyway.

 

Thanks.

The easiest way out of this, if disk3 is still OK, would be to resync the new parity, unfortunately the Areca doesn't pass the standard ATA SMART info, if possible connect disk3 to one of the onboard SATA ports and download a SMART report.

If disk3 is really bad and you still have the old parity intact and there weren't any writes to the array since the parity upgrade, you can do a new config and use the invalid slot command to rebuild disk3, or if there's not an appropriate spare do it the manual way and only then you can use the parity swap procedure.

I can try another paritybuild.

Smart shows as OK. Connected the disk to a different computer and checked.

It was extremely slow last time I tried it though. Like 16MB/s slow.

Try another parity sync, if possible connect disk3 to a SATA controller, grab and post diags if more issues or is going very slow.

Ooof... after plugging disk3 into onboard SATA-controller, I got errors on SMART and that it has reallocated a sector.

Diags attached.

 

edit:

Something else that worries me, is that there have been writes to the array (not much) after removing the old parity disk.

Will it still be possible to use it to restore disk 3 if I get a new one?

 

Parity build is still slow, to put it mildly. It's jumping between 8 - 13 MB/s.

I've let it continue to build for now.

 

edit2:

Parity build seems to have picked up speed. It's now at ~50 - 60 MB/s.

"Warning - reported uncorrect is 60" (SMART health)

"Warning - reallocated sector ct is 712" (SMART health)

 

smarterr-diagnostics-20181202-1820.zip

Edited December 2, 20187 yr by DogeKitteh007

Disk3 is failing, but there aren't any read errors during the sync, though it only covers the first 10 minutes, if there are no read errors let it continue, and even if there are some it's likely your best bet, that or using ddrescue.

 

6 hours ago, DogeKitteh007 said:

Something else that worries me, is that there have been writes to the array (not much) after removing the old parity disk.

Any writes to the array after removing the old parity make the other scenario not recommended, since you'd get some or a lot of corruption on the rebuilt disk.

Hopeful!

 

The parity build seems to have completed without errors.

Is it ok to pull the bad disk (disk3) and replace it with the new 8TB?

 

I've taken the array offline for now.

unraid-diagnostics-20181204-0220.zip

6 hours ago, DogeKitteh007 said:

s it ok to pull the bad disk (disk3) and replace it with the new 8TB?

Yes

Array is up and running again. Thanks for helping guys.