Buying a Seagate product increases failure rate exponentially!

But you have twice as many drives, so that doubles the risk of drive failure...

Potatoes potatoes.

Agreed, you should buy 2 8TBs.

I can't fault your logic. There is the same chance of one of these failing as an 8TB drive - even though there are 5 and a half million of them! I agree to an extent that a new 6/7/8-platter drive will contain some new and untested technology and be at a greater risk of failure than a single platter drive, but all of my 2TB drives contain 4 or 5 platters anyway. I'm not at all convinced that an appreciable difference exists. (4 or 5 platters would have been new and untested technology when I bought them). Taking an extreme approach (as I have with the floppys), I would prefer to have fewer drives of larger capacity. My old 500GB drives are now almost useless to me for backing up films, and the same will be true of my 2TB's much sooner than my 4TB's.

is this 2x4tb drive or 1x 8tb?
cheers

8TB is all well & good, but if it goes wrong, that's 8TB of potentially lost data. 2 x 4TB halves the risk.

True

But going with that logic, you'd have the same chance of failure, but only lose half the data.

I'm sure I've read that the larger drives fail more frequently anyway, so multiple smaller ones might be preferable to both reduce chance of total failure and bring overall failure rate down as well.

Buying a Seagate product increases failure rate exponentially!

Perhaps, but
http://www.digitaltrends.com/computing/backblaze-tests-seagate-8tb-drives-reliable/#ixzz4bsuAGLxv

I can't fault your logic. There is the same chance of one of these failing as an 8TB drive - even though there are 5 and a half million of them! I agree to an extent that a new 6/7/8-platter drive will contain some new and untested technology and be at a greater risk of failure than a single platter drive, but all of my 2TB drives contain 4 or 5 platters anyway. I'm not at all convinced that an appreciable difference exists. (4 or 5 platters would have been new and untested technology when I bought them). Taking an extreme approach (as I have with the floppys), I would prefer to have fewer drives of larger capacity. My old 500GB drives are now almost useless to me for backing up films, and the same will be true of my 2TB's much sooner than my 4TB's.

I bet you are one of those people who still think Duracell batteries are 100x better than all other brands :laughing:

In tests 8 out of 10 kittens recommend this drive

Always buy x2. Never put all your eggs in one basket.

Unless you already have 8TB's of HDD data awaiting a backup clone. Everything in duplicate is just so much easier than bothering with parity rebuilds.

Heat from me

Important documents, photos and videos backup with Google drive.

Everything else is replaceable to me...

Depending on the configuration you could have double the chance of a failure.

I always advise a proper backup regardless of the storage size, your point may have some merit, however nothing can help you like a backup!...

I watched all those adverts with the bunnies running about - you mean that they are not twice as good because they are twice the price? ~I want my money back

is this easy to open to remove the drive to use internally? i think the last 8tb drives i got from WD were about £180 with a 10% discount code. external drives but easily to remove for internal use

supposedly not.
a) you have to break in
b) the internal drive's warranty knows it was part of a consumer unit, rather than standalone HDD

I would love one of these... still a bit expensive for me tho, despite being a fantastic deal. Heat added.

Stats... tempting... me... arg, I'll explain.
If one drive has probability p (say 5%) to fail then you have these probabilities:
1) both drives fail together almost never with p*p = 0.25%. This is why mirror or raid makes sense
2) one of two drives fail with 2*p*(1-p) = 9.5%. You almost double your chance that one fails, because you have two in your basket now. This is why striping (RAID 0) is a bad idea without very frequent backups or other fault tolerance.
3) The probability of one or more drives failing (1 and 2 together) is 1-(1-p)^2 = 9.75%. If you had n drives this would be 1-(1-p)^n -> 100% as n grows. Failure is a certainty if you have too many drives...

And 1) 2) 3) together is the reason for RAID for data centers with large storage - they have failures every day and couldn't guarantee uptime otherwise.

For those thinking about removing this drive from the enclosure for alternative uses, it is worth noting that this drive is a Seagate Archive HDD (ST8000AS0002) that uses Shingled Magnetic Recording (SMR) technology.

If you want a NAS drive, I think this other Seagate, at £229, is a better purchase.

https://www.amazon.co.uk/d/Computers-Accessories/IRONWOLF-8TB-NAS-5IN-7200RPM-SATA-256MB/B01M1BUBSO

My worries with increased failure rate do seem unfounded after looking at backblaze data.

Thanks for doing some maths. I had originally though that the increased size would also increase failure rate, so I thought it was a no brainer. It turns out they might be more reliable, so some proper maths/stats is good to see. Lessened my worries about getting the bigger drives sometime soon, but not got the money or need just yet.

Glad you found my math interesting. I think it nicely shows that RAID can make sense if you have many drives and how important backups are.

I don't really know if newer, larger drives have a worse failure rate than the smaller ones. But I doubt the failure rate is massively different.

I'm not saying SMR disks are all bad... I have a one of these Seagate Archive drives in my microserver, but I use it as an archive drive (i.e. as intended). However, I would never use one of these in my main desktop, because I think performance is poor compared to other drives.

However, based on current pricing, I would spend the extra money to get a Seagate Ironwolf if I was buying another 8TB today for my microserver.

I agree, I don't use anything less than an SSD for my desktop - but you specifically recommended a drive for NAS usage, and it's pointless unless you have 10Gbps connection at home, even then, I struggle to believe anyone has content that needs a faster read speed than 1Gbps (at home).

These drives are no more or less reliable as storage disks than a WD Red

The read speed is good for the Seagate Archive drives, but the write speeds really slow down after the first few minutes, which is attributed to the fact this is a shingled drive. It is great for dropping music and movies, etc, onto and then only reading, but I also do daily and weekly backups from my desktop to my server (running WIndows Home Server), so I also want a drive with good write speeds... if the Archive drive and Ironwolf are similar prices, I see no good reason for anyone to pick the Archive drive.

Ah fair enough, I use an SSD as a staging area for larger files like system images or uncompressed rips etc., my system automatically moves files to the HDD after being copied to the SSD, so write speed isn't an issue for me.

How to get it £194.88. I ti s shown £209.99 on Amazon page

Received the drive today, all well so far. Passes in SMART tests in HD Tune Pro. Comes out of the box with high raw read/seek error rate values, but this seems to be ignorable on Seagate drives. Pic: https://s2.postimg.org/s6s59tvi1/drive.png