How bad is it and how much time do I potentially have.

MrSimplicity · January 22

So I have a HDD (Seagate IronWolf Pro 16TB) that was actually a replacement for a RMA'd HDD. However I'm almost certain this one is on the brink of failing and it's "warranty" is expired. These drives are $260 a piece and I didn't even have them collectively for a year. Some of my other drives get a couple errors here and there, two to be exact and one has 24 and one has 8 RSC) but that's after running for 6+ years continuously. It could be slightly higher but as far as I'm concerned I haven't even written a terabyte onto the drive yet. Out of the 4 that I originally bought 3 of them failed within a month and I sent them in. took me 3 months to get the first 2 back and another month to get the third. I guess not completely relevant but some fun context. Luckily though I'm on unraid and can easily move the data from one drive to another and just pop in a new one. Though not really trying to spend the money on another drive especially within a year of buying the last set of them. Also just because I mentioned it... Seagate support sucks. So how bad is it and is there a way to save the drive or should I just take the loss and swap it out. I could honestly just pull it and it wouldn't really hurt me too bad. I've only used 13 of the 80 TB in my NAS. So losing 16 for a while would be fine.

kokosnh · January 22

The surface is damaged, hard to tell without the surface latency test…

I would not trust my data on it.

this is just tu much work to get anything from this

Kilrah · January 22

Yeah this warrants replacement now.

Also Ironwolf Pros should have 5 year warranty so you should be able to use that.

But if you had 2 fail I'd suspect something environment-related like improper mounting/vibration isolation.

Also it's not recommended to add drives to an Unraid array until you actually need them, every drive becomes a liability and removing one is not trivial. When one drive fails all the others need to be in fully working order for a successful recovery, and it'd be very stupid to be unable to recover a drive because another that was unused failed during the rebuild.

MrSimplicity · January 22

1 hour ago, Kilrah said:

But if you had 2 fail I'd suspect something environment-related like improper mounting/vibration isolation.

All 8 drives are on drive sleds and there's no outside forces that would cause my server to shake or vibrate other than just the other drives.

1 hour ago, Kilrah said:

Also Ironwolf Pros should have 5 year warranty so you should be able to use that.

When I registered them on Seagate they only gave me like 3 months. Plus the drive that's failing now is a refurbished drive.

1 hour ago, Kilrah said:

When one drive fails all the others need to be in fully working order for a successful recovery

Not quite, I can just move the data between drives and then just take it out. My server is already idiot proof. If I lose any data, though I still have parity, it will see that and go get (download) it again.

3 hours ago, kokosnh said:

surface latency test

Is that something I can run on my server? Or else I can take it out and pop it into my main PC if I have to. Otherwise do you think just trash it and replace it?

kokosnh · January 23

13 minutes ago, MrSimplicity said:

Is that something I can run on my server? Or else I can take it out and pop it into my main PC if I have to. Otherwise do you think just trash it and replace it?

Best to use bootable program, like mhdd, or Victoria, diskgenius can do it too. But just from the smart data, there’s no point, to much damage already. and it would take lots of time, like a day or more with so much bad sectors.