|
There is at least one site that uses massive numbers of hard drives (thousands at least) and have been doing so for years.
They collect failures stats for all of those at an individual basis and make the results public. They have been doing that for quite some time.
Lots of detail.
Fun random read. And probably more relevant it you really need to get some real world stats on failure rates.
|
|
|
|
|
It's BackBlaze (2023 stats)[^]
I have looked at it in the past, but as they've moved to larger enterprise drives it seems less relevant - I don't need 12Tb or larger drives for home use! And, if I look at the stats by manufacturer, it's difficult to determine if there's any consistency or if it's down to specific models.
|
|
|
|
|
I believe Wayback Machine publishes (possibly 'published') similar data. Several years ago, I saw their figures; I believe it was on a web page. I worked on a collaborative project with them, so it could be that the data was not openly available to everyone. (But then: Why should it not be?)
I wouldn't draw too fast conclusions from one such log, though. Lots of factors may affect perceived 'disk quality'. Did you monitor the voltage stability of the power delivered to all of the disks, at the disk power contact, or could it possibly vary from one disk to another? What about that thunderstorm causing a lot of strong electrical fields - was the failure rate identical for the disks installed a week before the thunderstorm and a those installed a week after it? Do all disks have the same protection against environmental hazards such as temperature and vibrations? Are you sure that the truck delivering one batch of disks did not hit a bump in the road, giving that batch a mechanical shock even before they were installed?
Some disks have consistently high failure rates across a lot of installations. (I will not mention any specific manufacturer - It's Better Muting that ...). It would be a good thing if some web site would collect such statistics from a number of different huge installations, to show which manufacturers / models are consistently good or bad.
Religious freedom is the freedom to say that two plus two make five.
|
|
|
|
|
I read an article posted somewhere by google. Most of their data centers use SSDs but before that spinners.
That said, from a single developer to small develop shop, after 5+ years, new SSDs are ordered. Existing SSDs are cloned and tossed. Screwing around with old tech is interesting but not profitable. True story - had a customer 3 degrees away from me. They had spinners in backup storage - sealed in the event of active hardware failure. Some of these would not spin up, so they hit them with a hair dryer... here's your sign...
Charlie Gilley
“They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759
Has never been more appropriate.
|
|
|
|
|
dandy72 wrote: Also - I'd never buy a refurbished drive. Ask yourself: What are the reasons anyone would ever send a hard drive back?
I would never buy even a new a drive from Amazon. I thought to save some money and bought 2 8TB drives from them (in 2022, IIRC). Both were dead on arrival. Buying the drives locally in Israel is a bit more expensive, but none of them have ever arrived dead.
(Sample of >30 drives over the last 20 years)
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
I've only been buying drives from Amazon for...a decade, maybe? At least 20 drives. None of them have had a hiccup so far (and FWIW, all WD Elements or WD My Book).
|
|
|
|
|
StarNamer@work wrote: I'm wondering what his experience has been with them so far and how it matches with mine
So, I guess not knowing is driving you crazy. Yeah I know, I'll get my coat. 
StarNamer@work wrote: I'm beginning to think this may have been false economy and I'd have done better sticking with newer Seagate drives despite the price!
Keep in mind, I haven't built a RAID in years. So, my experience is old and crusty, but back in the day Seagates were always known to fail before drives like from WD. I'm sure someone online will get upset and emotional about that, but whatever. Anyway, what time span are you talking here? Assume the warranty was still valid then did all these drive failures happen within a year? If so, that's crazy.
Also assuming prior to this HGST fiasco you didn't have drives failing like popcorn popping in the microwave, which would indicate a problem with your housing... maybe it overheats (which is a big problem), or you're setting your enclosure on top of a large speaker magnet for funzies, etc... Then you might be onto something.
If HGST was acquired by WD in 2012 then it's safe to assume they it was acquired with inventory. Given the fact that we had serious economic trouble in 2007-2008 and inventory can be manufactured a few years before it's actually sold to the customer (depends on the size of the company), or even if manufactured in 2012 maybe they started in 2008 being cheap and continued with it. So, in theory it's possible there was some cheap batches made you were unlucky to get. It's just conjecture though. Either way, might be time to try a different brand.
Side note, Google used to keep a list of which drive brands suck. They go through millions of them and they know. They refused to release that list though as it would effectively put that company out of business... even though in a real free market that can and should happen. Which is to say, the consumer is the last person companies care about these days.
Jeremy Falcon
modified yesterday.
|
|
|
|
|
Jeremy Falcon wrote: Side note, Google used to keep a list of which drive brands suck.
Backblaze isn't shying away from that, and my conclusions seem to match their yearly reports.
RE: HGST...weren't they the ones that had a major flood at their manufacturing plant a decade+ ago, and subsequently had a huge batch of unreliable drives?
|
|
|
|
|
dandy72 wrote: Backblaze isn't shying away from that, and my conclusions seem to match their yearly reports.
Respect.
dandy72 wrote: RE: HGST...weren't they the ones that had a major flood at their manufacturing plant a decade+ ago, and subsequently had a huge batch of unreliable drives?
Dunno. I've been doing cloud everything lately, so me old and crusty with that. Never even heard of HGST until this post. It would explain a lot though.
Jeremy Falcon
|
|
|
|
|
I think HGST used to be a Hitachi brand, and I've purchased external drives from WD, and the drives inside had an HGST label. Even recent ones, so even though they might no longer promote the HGST brand (at least on the box), WD is clearly still using the name internally...
And I don't think I've had any sort of bad failure rate with the drives I have that I know are HGST.
|
|
|
|
|
They were and Hitachi bought that division from IBM earlier. The flooding happened in Thailand if I recall correctly.
"They have a consciousness, they have a life, they have a soul! Damn you! Let the rabbits wear glasses! Save our brothers! Can I get an amen?"
|
|
|
|
|
Weren't the IBM drives originally branded as "deskstar" which, of course, became corrupted to death star?
I've had reasonable experiences with Hitachi and Western Digital drives - no unexpected failures before they were replaced because they were either getting old anyway, or were just replaced for more capacity.
I did have to laugh when many years ago, having just bought and installed a Fujitsu Robin drive a "very knowledgeable" friend (a technical writer journo) posted a list of drives not to be touched with a dirty stick online, and the Fujitsu was top of the stack. That drive eventually got retired some 6 years later without a hiccup when the machine it was in was upgraded.
|
|
|
|
|
Alister Morton wrote: Weren't the IBM drives originally branded as "deskstar" which, of course, became corrupted to death star?
They were. I had the misfortune of buying a couple of these, and they are the only drives that catastrophically failed in service for me.
Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.
|
|
|
|
|
If you deep doin whatcha been doin, you'll keep gettin whatcha been gettin.
Definition of a burocrate; Delegate, Take Credit, shift blame.
PartsBin an Electronics Part Organizer - Release Version 1.3.1 JaxCoder.com
Latest Article: EventAggregator
|
|
|
|
|
In a closed society where everybody's guilty, the only crime is getting caught. In a world of thieves, the only final sin is stupidity. - Hunter S Thompson - RIP
|
|
|
|
|
This message has been flagged as potential spam and is awaiting moderation
|
|
|
|
|
Or a crackpot idea.
MSN[^]
|
|
|
|
|
|
It's a crackpot idea, don't get sucked in.
Definition of a burocrate; Delegate, Take Credit, shift blame.
PartsBin an Electronics Part Organizer - Release Version 1.3.1 JaxCoder.com
Latest Article: EventAggregator
|
|
|
|
|
|
Stop it with link only posts. (grumpy max)
At least copy the title or the summary if there is one.
CI/CD = Continuous Impediment/Continuous Despair
|
|
|
|
|
MSN[^] 18 Reasons Why Men Get Grumpier As They Age
|
|
|
|
|
|
touché!
CI/CD = Continuous Impediment/Continuous Despair
|
|
|
|
|
2 things:
(a) I see MSN still hasn't lost their habit of spreading their "[N] reasons why XYZ" type of articles over [N+M] pages (where M are just extra ad pages). So make that 19 reasons.
(b) Was there anything in there that was actually unique to men, and could not equally apply to women? Besides, aren't we're all supposed to be gender-fluid now?
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
