• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Hard Drive Product Lines & Reliability Factors in 2022

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Max0r

Member
Joined
Oct 18, 2005
Location
Chicago Burbs
All of PC part selection is about dealing with probabilities rather than certainties, but when it comes to most components, this tends to not be at the forefront of the mind, as failure tends to be neatly tucked away from the experience of most, hiding behind low percentages and high lifespans. But I dare say no component selection brings the reality of probability over certainty closer to the conscious mind than that of long term storage devices, namely HDDs. We know they're going to fail. We expect them to fail. And yet they offer the greatest balance between long term data stability and performance/utility/rewritability, between SSDs and tape/stone-tablets. Despite SSDs seemingly surging far ahead in supposed lifespan, they are dependent on being powered on much more often and maintained by their background operations. HDDs can go much longer powered off without invoking the anxiety of disappearing data and they are still bigger and cheaper. And in a way, the beauty of HDDs isn't that they don't fail, but that they tend to fail more reliably, if that makes any sense.

I ask of your opinions on this matter knowing full well that we're all playing the worst of guessing games, and yet somehow by attempting to gain a probabilistic edge I believe many of us have fared far better in the HDD department than our peers.

Long ago I was a die hard Western Digital Caviar/Black whatever fan. Several years ago with my previous build these notions were shattered. I was seeing an overwhelming torrent of negative reviews, especially on the cheaper "Blue" models, but even tons on the black models. Seagate Barracudas, another cheapo line, had terrible reviews as well, much like the Blue series. At the same time, HDDs were beginning to be categorized into different use scenarios, which, from a performance perspective, were absolutely laughable (save perhaps for very heavy server specific things), but from a potential reliability perspective quite intriguing indeed!

At the time I got the impression that drives marketed for surveillance and NAS uses were a cut above the fray. Not necessarily great, but better, as they'd both be, supposedly, made to withstand far more usage. "Black" type drives were basically a joke. Who needed "performance optimization" when you have SSDs or real enterprise/server type drives? To me they seemed like just a slight reliability upgrade from the Blue series type drives which to me almost seemed like an unholy bargain for those who just could quite afford SSDs or who were fine with the prospect of losing mass data. It would just be games or other BS anyway, right? It's not like you buy some Blue drives so you can actually run some kind of data preservation scheme. It's just a shitload of cheap storage that might last a long time but has a high likelihood of not making it past a couple years. It gets the job done, for a time anyway.

At the time I figured the best prospects were enterprise, surveillance and NAS drives. I had lost faith in any manufacturer and instead focused more on product lines and reviews. At the time Seagate Skyhawk Surveillance drives seemed alright looking so I got two of those. One as a normal drive, one as a backup. They've accumulated around 3 and something years run time so far without a hiccup. Later on I needed more space and was seeing very good things about the HGST ultrastars, and found a couple unbelievably cheap ones, well past obsolescence, but 3 TB each, supposedly new (many people say goharddrive just cleans the SMART data and pretends the drives are new, but others say it's not that simple and they are actually a good actor). Once again I did the same thing, using one as a backup for the other. Meanwhile, for the most important data I would regularly back it up to 3-4 of the drives, dropbox, and also have manual backup layers incase I accidentally batch screw something up. These two drives have accumulated so far about 1.5 years of run time, again without a hiccup.

It needs not be mentioned my experiences here are purely anecdotal; statistically insignificant. Aside from drive selection, it's been my impression that the best ways to maintain a drive's longevity are feeding it clean power, keeping them from heating up too much, keeping them off as much as possible but not switching them on and off too much either, and not hammering them constantly with ridiculous amounts of fragmentation and no free space to deal with, as if a VMEM swap drive without enough RAM.

Coming upon 2022 and my new build, in the interests of maintaining my progression of fresh long term storage space directly attached to my new system, and relegating my older 4 HDDs (not to mention my many others) toward mostly 24/7 power off tertiary backup status (occasional updates over the LAN, or pulling stuff off them notwithstanding), I have once again arrived at the ultimate dilemma of trying to optimize my HDD choice of maximum reliability prospects with decent price. One of the things that drew me to the Skyhawk Surveillance drives was not just what, at the time, were relatively favorable reviews, and the fact that they were "surveillance" drives, but also because they brought all this at a much lower pricepoint per GB than, say, WD purples did. They were significantly more expensive than the budget Blues and Barracudas, but significantly cheaper than many NAS/enterprise/surveillance product lines, while in some cases having comparable or superior reviews (albeit review quantity was rather thin, and suspiciously could have been solicited, which I accepted at the time).

And so today I've seen that any HDD model that showed promise in a backblaze test is instantly set upon by the hordes of the hard disk hungry, and all that remains is dodgy sellers pretending to sell new drives that have actually seen more action than your proverbial mother has. All that's left for affordability without being complete and utter, obvious trash, is refurbs, recertifieds, and suspicious products/sellers. When sorting by "rating" at pc parts picker, the worst product lines are at the top, despite seeing an exponential increase in % 1 star ratings. You see a bunch of 5's, 4's, then nothing at 3-2, and then 12% 1 stars of people, likely newly blackpilled on the realities of bargain basement $ per gigabyte product lines from manufacturers who have long since abandoned even the pretense of maintaining a reputation for quality.

And so I ask, to no one, hearing nothing but my echo... what are your opinions on finding the ultimate balance of price per gigabyte and reliability prospects in today's hard drive environment? And as an old former friend who is the one man responsible for getting me into the hardware level of computing so long ago said of one man turning to another in a steam room: "So... how big is YOUR hard drive?"
 
If I were to be completely honest, I haven't looked into HDD anything in years. Since moving to SSDs, my HDDs are merely storage vessels that generally hold cold data anyway so the speed is largely irrelevant (for me). I also have multiple backup locations (drives) so reliability isn't too much of a concern.

As far as reliability goes, I've taken a gander at the Backblaze info in the past. While neat, I'm not entirely sure data center reliability gives us an idea of real-world performance. You can also glean from that data the difference between (most) drives is small anyway. I mean you never know when a drive is going to crap out. Some early, some late...

In short, I just buy a HDD from a major brand and never looked back hoping not to run into the likes of a 'Hitachi Deathstar', lol. Feels a bit Iike playing in the minutia for HDD because the differences between most is so small... but maybe that's just me? All I look for is what the line is for (be it speed, efficiency, NAS, surveillance/recording, etc) and buy one that's a reasonable price.

While you may think people have done better or worse, everyone has a great, or horror story about (name a piece of hardware). But just because Earthdog's HDD crapped out early, doesn't mean yours will.... or it could die out earlier... The point is that we don't know. Would you not buy a drive that showed high reliability from Backblaze but a member or two here said it crapped out early? It doesn't make sense to use small data set info. I mean, you're more likely to flip a coin on heads 5 times in a row than 20........ at least this is the way I look at it these days for how I use HDDs at home.
 
In the last 2 years, the only HDDs that failed at work (various clients) were Seagate drives that were working in NAS for 4-5 years 24/7. We have a lot of clients with 2-8 HDD NAS and most drives there are from WD, but still Seagate fails much faster.
I noticed that WD improved the Red series, so it's not failing like its early version. There is Red Pro released a while ago, also improved recently. WD Ultrastar is recommended by network storage manufacturers. Seagate released their NAS series a while ago and it replaced the whole old series, so it's hard to make stats based on HDD which are working for a couple of months.
HDDs are used mainly in companies right now. Home users are moving to SSD for a couple of years already and most users don't need higher capacity storage at home. Most laptops don't have HDD. Most home/office PCs are not sold with HDD anymore. If you are using HDD for a backup then you always use more than 1 drive and RAID1 or higher option. More HDD series are sold with a 5-year warranty nowadays than some years ago.
I just can't see the problem here.
 
In the last 2 years, the only HDDs that failed at work (various clients) were Seagate drives that were working in NAS for 4-5 years 24/7. We have a lot of clients with 2-8 HDD NAS and most drives there are from WD, but still Seagate fails much faster.
I noticed that WD improved the Red series, so it's not failing like its early version. There is Red Pro released a while ago, also improved recently. WD Ultrastar is recommended by network storage manufacturers. Seagate released their NAS series a while ago and it replaced the whole old series, so it's hard to make stats based on HDD which are working for a couple of months.
HDDs are used mainly in companies right now. Home users are moving to SSD for a couple of years already and most users don't need higher capacity storage at home. Most laptops don't have HDD. Most home/office PCs are not sold with HDD anymore. If you are using HDD for a backup then you always use more than 1 drive and RAID1 or higher option. More HDD series are sold with a 5-year warranty nowadays than some years ago.
I just can't see the problem here.

Fortunately I don't give a damn about new product lines, just old ones that have had time to be put through the ringer already. By-the-by, you wouldn't happen to know what product lines those Seagate drives were from would you? Were there any surveillance drives? Just NAS drives? Any of the cheapo Barracudas?

Whatever happens, I'm probably just going to cruise the NAS/enterprise/surveillance listings and look for the best balance of price/GB and reviews. Most likely in the 3-4 TB range. The sting of 1st world poverty is real, not being able to run a 3-drive failure tolerant file server ^_^
 
My philosophy behind storage is simply put, this:
NVMe for the OS and oft-used software like MS Office and a few games.
A larger NVMe for general storage, short term. (E.G. Client files that I delete after a month if the client doesn't ask for anything.)

My long term storage is on spinners. RAID 5 is my preferred configuration. Right now I use UnRAID which allows me to upgrade my storage to larger sizes one at a time. It is in a RAID 5 type config in that I have a parity drive (equal to or greater in size to ANY of the other drives).

With that in mind, the branding means little to me. I buy cheap and replace as needed. With parity in my long term storage, I'm safer than the average bear. I once followed the "Friends don't let friends buy Seagate" but times change and that saying has aged like milk. Maxtors were also known to fail often but they were the speed kings back in the day.

E.D. The "Deathstar" was IBM when it got that nickname and Hitachi bought them lock, stock and barrel. ;)
 
Whatever happens, I'm probably just going to cruise the NAS/enterprise/surveillance listings and look for the best balance of price/GB and reviews.
If that's what you're using them for, sounds like a plan.
 
I've been using spinner drives for storage for ages. I've had a few failures (1 Western Digital, and several Toshibas). Toshiba drives are the worse. They drop like flies and I stopped using them. Currently using a pair of Western Digital Reds in raid-0 for games and a pair of WD Velociraptors in raid-0 for media. Tho my spinners are over a decade old, they are still running strong and are very reliable. I'll keep using spinner drives for storage as they're cheap. As a matter of fact I just snagged a pair of Seagate Ironwolfs 6tb to replace the WD Reds 3tb as I'm running out of space for my games.

Nvme drive as main for OS/wares. What I'm disliking of Nvme drives is their life expectancy is dwindling. I wished they lasted as long as mechanical drives.
 
So after a lot of shopping around... seeing the only affordable hard drives being the worst of the worst, realizing almost any "new" old model Ultrastar is a scam unless it's $100+ or SAS, after noticing that 12 different sellers on Ebay were offering "new" Ultrastars for only $56-59, and yet mysteriously, every single seller was located in "Israel" and almost every one was shipping from Arizona. They all had lots of positive feedback, and yet, this seemed just too shady. All their listings looked exactly the same of course.

I just broke down and decided to take a chance and try something I've never done before: Used hard drives. I bought a couple "factory recertified" ultrastar 3TBs from Goharddrive. They actually have a few different 3TB ultrastar listings all for $50 each, all used. But all have 1 year warranties except these. These are the only ones that say "factory recertified" and have a 3 year warranty from the reseller. I've also seen endless comments from hardcore storage array builders that they've had very good experience with Goharddrive warranties and many of those ultrastars they bought used lasted years of run time thereafter and still not dead. Many say that a used ultrastar is better than most new hard drives. That being said, I've also heard stuff about Goharddrive pretending to sell new drives but just selling wiped SMART drives. One of their $50 ultrastar listings does not say anywhere that it's refurbished or recertified, but if you scroll down the condition says "OEM," I think they had an oversight there. And it only has a 1 year warranty. If they are trying to sell ultrastars as new I didn't scroll high enough in their pricepoints to see them.

The ultrastars I bought from Goharddrive a few years ago were labelled as new, and came with a 5 year warranty. I also got them for $50 at the time which IMO is dirt cheap for a new one. They appeared new to me on the SMART, but I'm entry level when it comes to looking at that stuff, so they could have easily fooled me. But if Goharddrive is good about their warranties, then I would say you can easily guess the overall newness or relative quality of refurb based on the warranty period they offer. So I think 3 years is pretty good.

We'll have to see. I will say this. IMO a new WD Blue or Seagate Barracuda is more of a liability than a used Ultrastar.
 
I think the differences between then isn't much, overall. One might have 2x the chance of dieing than another, but if the failure rate is less than 1% and then doubled...

You can have a 'good' drive die early and a 'poor' drive go long too. It's not my favorite dataset but look at the backblaze info..Ike woomack said, failure rates are all really low.

Edit: weird thought this was the WD thread.... same concepts, lol
 
Well, the hard drive game is a lot like day trading, or approaching a light at an intersection you KNOW has a 95% chance of turning yellow before you get there, but it FEELS like you can just squeeze past it if you give enough gas. To maximize your odds of consistent success, you have to maximize your statistical edge and do what is most likely to lead to the best results when repeated 1000 times. That means you do what your analysis has shown to give you the best combination of win-rate vs risk/reward when repeated over and over again. You can't control the outcome, and you know that there will always be dumb luck and failures. But you do what you know to be correct because you know that if you do that, it maximizes your odds for a successful outcome over a large number of iterations. That means in day trading, you follow the rules of your system no matter what, even if it means taking a loss sometimes, because overall your performance is most likely to be maximized by doing so. When you see that intersection approaching, no matter how tempting it is to keep on the gas instead of coasting to a stop, even though the light is green, you coast to a stop, because 95% of the time it WILL turn yellow before you get there. It also means, when you take in all the information about hard drives from experience, others' testing, reviews, etc, you make the best decision possible, when operating with limited funds, and limited time for bullshit after the purchase.

Yes, it might fail anyway, but if you don't play the odds according to strategy, you're not steering the ship in the direction of optimal outcomes.
 
Yes, it might fail anyway, but if you don't play the odds according to strategy, you're not steering the ship in the direction of optimal outcomes.
I agree here to a certain extent. For me at least, gone are the days of dredging details up for a 1% difference in anything (ok, I can count cards decently in blackjack....lol....I use that 1% difference, lol). I guess at heart I'm a gambler and take my chances. If you ask a forum and get enough opinions, it will all even out. But everyone has horror stories as much as success. The horror stories are pre-lubricated to come to life, however, remember that.

For storage purposes, I also have good backup and recovery plans for all of my data so if a drive fails, there's 'always 'another one with that same data. My most exposure is when I'm down to one drive (ordering another and getting the data on the dupe/letting the array rebuild). I find when adding time and effort in, just buying from what brands you know after a cursory search/review on the product, and go has worked well for my situation.

EDIT: But, I've said that much already, haha, sorry, lol!
 
Last edited:
I agree here to a certain extent. For me at least, gone are the days of dredging details up for a 1% difference in anything (ok, I can count cards decently in blackjack....lol....I use that 1% difference, lol).
Ahh, so you too understand the significance of playing your statistical edge. Btw, one guy I learned a lot about trading from goes around winning money from casinos in blackjack. He claims he's not counting cards, but clearly he's implementing some kind of special strategy. He even @'s the casinos in tweets about it, so apparently he isn't afraid of being banned. And they don't seem to mind much either. I think he's full of it. How is he not counting cards? But then again, he has a knack for coming up with simple but surprisingly effective trading strategies that I've never seen anyone else talk about. The greatest myth of all is that it's hard to find an edge. There are edges upon edges everywhere. The hard part is building a protocol to exploit them!
 
My strategy: SSD for O/S as others said. The only NVME I bought failed prematurely, was not able to obtain an RMA so I was a bit 'put-off'.
Using a WD Purple for local storage on one main desktop, a Seagate 5400 RPM,(4TB for $69) for desktop #2. Plus a couple externals for additional redundancy.
 
Back