C: Primary Boot HD: WD_BLACK SN770 2TB (2Tbyte NVME)
D: RAID1 (data only) SATA HDs:
- #1 *WDC WD4000FYYZ-05UL1B0 (4 Tbyte)* <--- ***BAD HD***
- #2 ST4000NM0035-1V4107 (4 Tbyte)

Not Assigned a Drive Letter: TEAM TM8FP6128G (Used to dual-boot into Linux Mint)

I built this PC 2 years ago, and set-up two 4-Gbyte HDs as a RAID 1, and now the software says the RAID is "Critical".
The research I've done says I need to replace the defective HD with a good one, BUT

In BIOS I turned off the RAID and selected AHCI, and then ran Victoria on both HDs. Victoria finds 16 uncorrectable blocks on the WDC 4 Tbyte HD,
but I want to know that it's 100% certain that the HD cannot be repaired and put back into use, before replacing it.

Also, I've never replaced a bad HD in a RAID1, and am afraid I'll do something and destroy the data that is on the good HD, so before doing
anything else, I'm here to find out what I should do.

I wonder if the common-sense 1st step is to simply indicate that the "bad" HD is a "spare" (whatever the RAIDXpert2 language is), and just let the
RAID software try to rebuild the array. Maybe the 16 uncorrectable blocks can be remapped? And again, I'm worried about doing something that
gets the data on the good HD overwritten.

Thanks in advance.

Why did you run a high intensive I/O program on data you define as critical?

If the data is truly critical, treat it as such.

Create a 1:1 image / clone of the healthy (Seagate) 4TB drive using a program like OpenSuperClone. Then rebuild the array to a new drive by assigning the drive you replaced (WD) as a hot spare in the BIOS.

I didn't define it as "critical", RAIDXpert2 said it was. I have no idea what that means. I'm trying to avoid buying a 3rd 4Tbyte HD. I want to know if the "bad one" is actually bad, and if it can be repaired, and the RAID rebuilt, and I want to know how to do it. I ran Victoria because it looked like "chkdsk" on steroids. Thought perhaps it would somehow fix things. Also the data isn't critical, but there is a lot of it, and I'd rather not lose it.

My bad - I incorrectly interpreted CRITICAL as your data, not the status of the array. Most controllers will simply say "Degraded" - though "Critical" sounds way cooler.

The actions you are performing will not silence the SMART warnings - so the output from your array will remain the same (in your case "critical"). The bad one is actually bad. That said, your next step is to perform a full SMART on the drive - so we can see all values; not just those which Victoria produced. This will likely provide proof of the claim.

How do I perform a "full SMART" on the drive? Special software?

How can you be certain the drive is 100% bad? In the past, I've run chkdsk and it does some kind of "repair", and things are fine after that.
I can't run chkdsk, or any other software that operates on a single drive, while it's configured as RAID, so I'm outside of my standard routine.
Isn't there any kind of whiz-bang software that can be run?
I'm actually prepared for purchasing a replacement HD, have it picked out on NewEgg, and if/when this WD HD gets reliably condemned, I intend to check here to see if the replacement HD is a good choice.

What I was mostly hoping for (here) was to find out how to indicate the good HD as the "original" and have the AMD RAID software (RAIDXpert2) just try to rebuild the array, but the main concern here was that I'd get it backwards and overwrite the good data on the good HD with whatever is on the bad HD. I get the idea that this software isn't very popular, as there isn't much documentation, and other people have let me know that software RAIDs are not good (for reasons I don't understand) and that if a person is determined to run a RAID 1, a dedicated hardware card is the way to go, so I have my eye on that idea as well.

This whole effort is, at least in part, a learning endeavor. I figured actually having a working RAID on my PC would be a good way to learn. And so, I have a concern because when I had RAID turned off in BIOS, and AHCI selected, and both HDs were being displayed by Win10's Disk Management as appearing normal, but for some reason I could not navigate either of them. I assume the RAID "did something" rendering the data inaccessible unless it's configured as a RAID, which is the exact opposite of what I thought would happen if one of the drives died.

So, I have @3 Tbytes of data on a "good" drive that I cannot access unless it's configured as a RAID. This is not what I expected, is the point. Next reboot I'll take a screen shot of what I'm looking at, and maybe you can explain what I'm seeing, and why.

Individual RAID members cannot be modified outside the RAID. If you run CHKDSK on a single drive which are members of a RAID you could corrupt your volume, I would not do that.

Running SMART (assume you don't R/W mount the drive) is fine. There are many tools that read SMART; given you've already connected the drives directly to Windows individually, I would suggest hddscan.com.

To answer your next Q:

CHKDSK = attempts to fix the file system.
SMART = health of disk internals, not file system / has nothing to do the filesystem.

SMART is telling you the internal functions of the drive are failing. Once this occurs, it cannot be "fixed".



I don't have to be certain, your hard drive is telling you its encountering R/W issues. Your interpretation of "good" or "bad" may be different than mine. For me, a good hard drive is one that does not have R/W issues.

Perhaps this will help: https://www.asus.com/ca-en/support/faq/1049870/

Full SMART from the Victoria SMART tab. Special programs are used to read RAID disks under Windows. You cannot write to this disk or run CheckDisk. The data will be corrupted.

RAID is passe which today is given over JBOD. RAID5 tolerates one failed disk while RAID6 tolerates two failed disks.

_________________
USB 3.5" enclosures are $19 USD postage paid from China, I own several