1 hour ago, ActionJackson said:

starting my array with the 3 drives

10 hours ago, trurl said:

You can New Config with nothing assigned as disk4 and let parity rebuild. Then you won't have a disk4 as part of the array.

But you will have valid parity that is in sync with those remaining disks.

1 hour ago, trurl said:

On mobile now will look at diagnostics soon to see why the errors for disk 2

For WD disks, we often recommend adding SMART attributes 1 and 200 to those monitored by Unraid. SMART attribute 1 for disk2:

  1 Raw_Read_Error_Rate     POSR-K   200   200   051    -    75077

It is over 9 years old and has never had a self-test. Syslog makes me think it is bad too.

Parity is over 10 years old, but it still looks OK. It also has never had a self-test.

Disk1, over 10 years old, but may be OK, never completed a self-test.

1 Raw_Read_Error_Rate POSR-K 200 200 051 - 2836

Disk3, slightly less than 10 years old, probably OK for now, never had a self-test.

1 Raw_Read_Error_Rate POSR-K 200 200 051 - 1

Just for completeness, "new" disk4, slightly more than 1 year old, never had a self-test but looks OK.

I'm not sure you will be able to complete a parity rebuild with disk2. It might be better to try to clone it to a new drive before trying to build parity.

Disk4 is probably the only one I would consider keeping.

But you can only reliably rebuild one at a time, and only if all remaining disks can be reliably read.

Sorry I didn't dig a little deeper into all your disks problems. That original disk4 I assumed was the main thing to be concerned with since it didn't really give a meaningful report.

And, usually, since each disk is an independent filesystem, one problem disk isn't going to cause problems with other disks. But some of the other disks have problems of their own.

I would say that all of your data is at risk, and normal Unraid activity may not be able to help that much.

It might be better to just forget about parity helping in any way for now, and just work with each disk individually with ddclone or even ddrescue to get their data onto new disks.

At the very least, I think that will have to be done with disk2 before you can consider building parity.

I am going to ask @JorgeB to look at this thread again. He hasn't posted to it for a while. But it may be a few hours before waking in his timezone.

I guess the good new is, you can start over with newer larger disks if we can just pull the data from those very old disks.

2 hours ago, ActionJackson said:

I have a 4TB I would like to install as a new parity drive, and depending on what is going on with that disk 2 I may need to replace it with my current 3TB parity drive.

I didn't specifically address this. If you had valid parity with only those remaining disks, then parity swap would let you copy parity to a new larger drive and rebuild disk2 to old parity. But you don't have valid parity with only those remaining disks.

ok, I installed a new(er) 4TB hard drive, did "new config", did not select a disk 4, did not select the new 4TB drive, and started the array so that parity would rebuild. In the morning, I will check it, and if all goes well I will swap parity drive to the 4TB drive, and put the old parity drive in my array.

I know these drives are old, and failing. Financially I'm not really able to do much about it.

26 minutes ago, ActionJackson said:

if all goes well I will swap parity drive to the 4TB drive, and put the old parity drive in my array.

Be sure to read the parity swap procedure link. That will at least allow you to avoid yet another parity build. It will copy parity onto a new disk and rebuild disk2 to old parity.

I will check back tomorrow.

Disk2 appears to be failing; the errors are logged as a disk problem, can run an extended SMART test. They can sometimes be intermittent, but my recommendation would be to replace that disk as soon as possible.

4 hours ago, JorgeB said:

replace that disk as soon as possible

10 hours ago, trurl said:

normal Unraid activity may not be able to help that much

Parity is not valid, so it can't be rebuilt.

I was thinking maybe ddclone it and then New Config with the clone to get parity rebuilt.

Wasn't a disk being rebuilt? It should continue even with a few read errors. Of course, it won't be 100% valid, but ddrescue will also skip the read errors, though it may recover a little more data from the bad sectors, so if parity is not valid, it may be the better option.

Original disk4 was unusable, but disk2 had already been disabled for some time. Disk4 was rebuilt to a "newish" disk, but corrupt, probably due to disk2 being very out-of-sync and also not completely readable. xfs_repair gave up on it.

Disk4 is being examined with UFS Explorer but could probably be put back in the array to rebuild disk2. I don't think the array has been started without disk4.

I removed disk4 from the array last night so that it could be scanned, and started the array without it so that parity could be rebuilt, which is now done. I plan on upgrading the parity drive to 4TB, replacing disk 2 with the current parity drive, and eventually moving disk4 back into the array once I've pulled the data off of the recovered drive, which is currently in progress

Post new diagnostics

I will post new diagnostics once the parity drive has been replaced, and disk 2 is removed. See you in a few hours

16 minutes ago, ActionJackson said:

I will post new diagnostics once the parity drive has been replaced, and disk 2 is removed. See you in a few hours

I am concerned that you are digging a deeper hole, trying to continue to work with that failing disk2 in the array.

Are you using the parity swap procedure as I mentioned?

Or are you instead trying to replace and rebuild parity?

46 minutes ago, ActionJackson said:

started the array without it so that parity could be rebuilt

And really no good reason to rebuild parity to that same disk just so you could turn around and create parity on another disk, whether you were planning to do parity swap or replace and rebuild. Just more wear and tear on these old disks and more likely corrupt parity due to the bad disk2.

1 hour ago, trurl said:

Disk4 is being examined with UFS Explorer but could probably be put back in the array to rebuild disk2. I don't think the array has been started without disk4.

It may be worth a try, as long as it's rebuilt to a new disk there's Nothing to lose; keep old disk2 intact for now, can always try ddrescue later on it.

Last night I did not realize I could do the parity replacement and a disk swap at the same time (sorry, this is all pretty new to me), and because everything takes like 6-8 hours I just started the array so parity would rebuild, and it seemed to.

Currently, Yes, I am doing the parity swap you mentioned, and replacing disk 2 at the same time so I can get it out of there. Once that's done, we should have 4 error free disks, albeit a bit on the old side.

Just now, ActionJackson said:

I am doing the parity swap you mentioned

Good!

That means bad disk2 will not even be used again.

Just now, trurl said:

That means bad disk2 will not even be used again.

Keep bad disk2 anyway in case it might be useful as an Unassigned Device to recover some files.

parity copying is half done, i know you're so eager Trurl :)

The copying is complete, but it says disk 2 is being emulated. The array is not on

storemore-diagnostics-20260311-2034.zip

my mistake i missed the part where it said to start the array. i started it and it looks like the new disk2 is re-building