Seagate highest failure rate?, why?

[ViruZ_HUN]

source: http://www.lifehacker.com.au/2014/01/the-most-and-least-reliable-hard-drive-brands/

 

So I see that seagate have the highest failure rate ratio?, lol but I heard that seagate have one of the best drives?, That website says that hitachi is best, I will do some research about Hitachi. So is it website tests true?

 

It says 1.5 terabyte drives have highest failure rate after 3 years. So it means that my two seagate 1.5 terabyte drives will soon die, lol

[ZetZet]

don't trust these. ever.

[qwertywarrior]

OLD NEWS

 

and turns out they were using hard drives from externalharddrives

opening it up and using them

 

they were banned from buying them too (from stores)

[helping]

there were MULTIPLE flaws in the testing environment

it's pretty much brand shaming attention whoring for that click click pesos

[dlf]

So plainly, yes they do fail more or not they do not fail more than Hitachi, WD, and other brands. My two external seagate drives (a 1 and 2 TB) do seem to get hotter than my WD.

[TheRetiredSlave]

I'm pretty sure that that article was blown wide open as being horribly flawed on so many levels..

 

Seagate is actually pretty good. (or at least my Seagate drives have all been pretty good)

[alpenwasser]

The methodology that was used to obtain that data is flawed

at best, and misleading at worst. Don't trust it.

No HDD manufacturer could keep themselves in business if

failure rates really were this high. Apparently Backblaze

even state that they'll continue to use Seagate drives,

which, to be honest, I sincerely doubt they would do if

they actually had any confidence in their own data (so,

they basically admit that the data is unreliable).

The only way to get reliable and statistically significant

data about this issue is to have all the drives running

under the exact same tightly controlled conditions (i.e.

properly using the scientific method, not just throwing

some random stats together) and then doing proper analysis

on the resulting data before releasing it.

Until I get such data, such "studies" don't really allow

to draw any reliable conclusions.

an example counter-article

EDIT: Too slow.

But yeah, as a few others have stated, don't trust that

article.

[squirrl]

Even if you did believe the article, and even with what you've written, it doesn't say that your drives WILL die after 3 years. It just says they have a higher chance of failure now than before.

Any mechanical drive as a higher chance of failure the older that it is, but it doesn't have to be an extremely high failure rate.

[T.Vengeance]

Anecdotally, 3 WD drives have failed on me. My 2 Seagates are still running fine.

[Ahnzh]

There are a few things coming to mind:

1. All HDDs were from one batch. It's commonly known that Seagate had a bad Barracuda batch.
2. Lately WD had a bad Reds batch as well with 40% failure rates on that batch.
3. Their testing method was flawed.
4. It's a long time test. Seagate acquired Samsungs HDD branch early 2012 and their failure rates got a lot better (they weren't bad in the first place). So the overhauled quality didn't even get into the test.
5. Seagate's Enterprise HDDs have a MBT of 1.400.000 Hours. that's 160 years...
6. It would be weird to have more failures in 1-3TB drives than 4TB drives.

[Ahnzh]

I really don't know when it happened that Seagate fell behind WD (in terms of reputation) but They always were comparable. I tend to buy Seagates, others prefer WDs. Hitachis were always better but one and a half times the price.

 

How did that reputation thing happen, does anybody know?

[flibberdipper]

Never trust these articles. EVER.

[IdeaStormer]

They all die, but I give you this we still have some 9GB SCSI Seagate drives running 24/7 for get this 15+ years! The 9GB SCSI should of been the give away on the age. Then I've seen new drives die after installing the OS, freaking drives, not often though but has happened. Still all drives are better now than before, less oddities even running with bad sectors is better than it was before, yea I'm not pulling a drive just for a few bad sectors it still works and the drive knows how to deal with it so it stays installed running along doing its thing.

 

Remember all man made devices have and will have defects and some will be way better than others, just get over it.

 

Forgot:

Those 1.5 TB drives had a firmware bug that if you updated it would give you a better drive, most never updated it.

[ViruZ_HUN]

The methodology that was used to obtain that data is flawed

at best, and misleading at worst. Don't trust it.

No HDD manufacturer could keep themselves in business if

failure rates really were this high. Apparently Backblaze

even state that they'll continue to use Seagate drives,

which, to be honest, I sincerely doubt they would do if

they actually had any confidence in their own data (so,

they basically admit that the data is unreliable).

The only way to get reliable and statistically significant

data about this issue is to have all the drives running

under the exact same tightly controlled conditions (i.e.

properly using the scientific method, not just throwing

some random stats together) and then doing proper analysis

on the resulting data before releasing it.

Until I get such data, such "studies" don't really allow

to draw any reliable conclusions.

an example counter-article

EDIT: Too slow.

But yeah, as a few others have stated, don't trust that

article.

 

 

There are a few things coming to mind:

1. All HDDs were from one batch. It's commonly known that Seagate had a bad Barracuda batch.
2. Lately WD had a bad Reds batch as well with 40% failure rates on that batch.
3. Their testing method was flawed.
4. It's a long time test. Seagate acquired Samsungs HDD branch early 2012 and their failure rates got a lot better (they weren't bad in the first place). So the overhauled quality didn't even get into the test.
5. Seagate's Enterprise HDDs have a MBT of 1.400.000 Hours. that's 160 years...
6. It would be weird to have more failures in 1-3TB drives than 4TB drives.

 

160 years?, lol seagate drives gets older then a Human:DD

They all die, but I give you this we still have some 9GB SCSI Seagate drives running 24/7 for get this 15+ years! The 9GB SCSI should of been the give away on the age. Then I've seen new drives die after installing the OS, freaking drives, not often though but has happened. Still all drives are better now than before, less oddities even running with bad sectors is better than it was before, yea I'm not pulling a drive just for a few bad sectors it still works and the drive knows how to deal with it so it stays installed running along doing its thing.

 

Remember all man made devices have and will have defects and some will be way better than others, just get over it.

 

Forgot:

Those 1.5 TB drives had a firmware bug that if you updated it would give you a better drive, most never updated it.

 

 

Thx all for the information, seems like this is a fake website, I got pretty suprised myself about the failure rates, I always thought that seagate was one of the best, and they still are since this is fake..

 

But guys i am still curious, Where can I see which is the mosr reliable drive brand?, The 2 brands i most trust it Hitachi and Seagate since I have Seagate and my Brother Hitachi.

 

My Seagate drives started to make strange sounds little bit then a year ago, My brothers Hitachi seems to have less sound, I dont know if this means something...

 

So which is the "best" HDD brand?

[IdeaStormer]

160 years?, lol seagate drives gets older then a Human:DD

 

 

Thx all for the information, seems like this is a fake website, I got pretty suprised myself about the failure rates, I always thought that seagate was one of the best, and they still are since this is fake..

 

But guys i am still curious, Where can I see which is the mosr reliable drive brand?, The 2 brands i most trust it Hitachi and Seagate since I have Seagate and my Brother Hitachi.

 

My Seagate drives started to make strange sounds little bit then a year ago, My brothers Hitachi seems to have less sound, I dont know if this means something...

 

So which is the "best" HDD brand?

 

They are all good, the best is usually based on which one a person owns and is still running so they mark it "the best" and tell the world its the best. Then some schlub buys the same drive and its DOA so they scream at the top of their lungs that that brand is the worse! Of course, if you want something that is theoretically better buy one based on the warranty length, of course the longer the warranty the more you'll pay, those are also usually enterprise level drives, you don't get anything for free so don't start complaining that they cost more.

 

Personally I like Seagate, Hitachi and Fujitsu (yea Fujitsu), in the order of what they will be used for not by just the name, then I only go for enterprise level drives. I've seen them all die so don't think they are not without their problems but of them all WD has the worse case in my hands so I don't get them but have seen them roll for years like the other brands as well.

[alpenwasser]

I really don't know when it happened that Seagate fell behind WD (in terms of reputation) but They always were comparable. I tend to buy Seagates, others prefer WDs. Hitachis were always better but one and a half times the price.

 

How did that reputation thing happen, does anybody know?

 

I think there was a Barracuda generation which had a lot of failures

a few years back, that kind of stuff tends to stick unfortunately

(I still see some people refer to Hitachi's drive as "Deathstars" due

to some bad experiences they had a while back with the brand).

Unfortunately I can't recall which generation it was or when. Personally

I used to buy mostly Samsung drives until they were bought out by Seagate,

since then I've mostly bought WDs (RE4s and Reds). So far neither of those

has let me down.

 

Thx all for the information, seems like this is a fake website, I got pretty suprised myself about the failure rates, I always thought that seagate was one of the best, and they still are since this is fake..

It's not so much fake as just unreliable I'd say. I'm assuming

their raw data is true, the problem is that when you wish to draw

meaningful conclusions based on that data, you need to take a lot

of other factors into account that just weren't documented in

this case (and maybe you need to revise your entire approach

for actually collecting the data, such as more tightly controlled

conditions).

As a very rough analogy: Take two teams with the same number of

workers and analyze their output over one month. You notice

that Team A has 50% higher output than Team B, in total.

Natural conclusion? Obviously Team B is lazy, right? Not so fast,

because half ot Team B's workers were sick for two weeks, Team A

has better equipment and because the remaining members of Team B

were overworked due to their collegue's absence there was an accident

taking another worker out of commission (and this is still leaving

out a ton of other mitigating factors, of course).

The simple statement that "Team A produced 50% more output than

team B" is pretty much useless as a basis for any decisions because

the underlying parameters of the situation are not taken into

account.

Similar thing happened here: There is just too much background

information missing for the summarizing statement to have any

meaning (i.e. it's not actually a reliable representation of

true drive reliability).

[Ahnzh]

 

-snip-

Your posts' formatting always seems like some kind of poem

 

'Deathstars' made me laugh hard LOL

 

Like i said, it was the same for me with samsung drives. Switched to WD as well but they let me down sadly...

[ViruZ_HUN]

 

I think there was a Barracuda generation which had a lot of failures

a few years back, that kind of stuff tends to stick unfortunately

(I still see some people refer to Hitachi's drive as "Deathstars" due

to some bad experiences they had a while back with the brand).

Unfortunately I can't recall which generation it was or when. Personally

I used to buy mostly Samsung drives until they were bought out by Seagate,

since then I've mostly bought WDs (RE4s and Reds). So far neither of those

has let me down.

 

It's not so much fake as just unreliable I'd say. I'm assuming

their raw data is true, the problem is that when you wish to draw

meaningful conclusions based on that data, you need to take a lot

of other factors into account that just weren't documented in

this case (and maybe you need to revise your entire approach

for actually collecting the data, such as more tightly controlled

conditions).

As a very rough analogy: Take two teams with the same number of

workers and analyze their output over one month. You notice

that Team A has 50% higher output than Team B, in total.

Natural conclusion? Obviously Team B is lazy, right? Not so fast,

because half ot Team B's workers were sick for two weeks, Team A

has better equipment and because the remaining members of Team B

were overworked due to their collegue's absence there was an accident

taking another worker out of commission (and this is still leaving

out a ton of other mitigating factors, of course).

The simple statement that "Team A produced 50% more output than

team B" is pretty much useless as a basis for any decisions because

the underlying parameters of the situation are not taken into

account.

Similar thing happened here: There is just too much background

information missing for the summarizing statement to have any

meaning (i.e. it's not actually a reliable representation of

true drive reliability).

 

 

It seems i will be looking into Hitachi drives if they are cheaper then seagates

 

When i buy my PC i will be looking into a 4 TB drive from some of them

 

I always seen WD as the priceier one

[ViruZ_HUN]

Also whats a NAS HDD?

 

whats the difference from a regular "desktop" HDD?

[alpenwasser]

It seems i will be looking into Hitachi drives if they are cheaper then seagates

 

When i buy my PC i will be looking into a 4 TB drive from some of them

 

I always seen WD as the priceier one

Yeah, definitely look at prices. Although where I live, Hitachi drives

are quite a bit more expensive than the others, although some of them

do have features that normal desktop drives do not (I recall some Hitachi

desktop drives being rated for 24/7 operation, for example).

 

Also whats a NAS HDD?

 

whats the difference from a regular "desktop" HDD?

It depends on the drive. Usually, NAS-suitable drives have different

error recovery behavior than regular drives.

@Vitalius explains this pretty well in this post, so I won't

type out an entire essay about it here.

Aside from that, whether or not they offer any tangible advantages

over regular drives is somewhat debatable in my personal opinion,

and depends on your specific usage scenario. In some cases they

might really offer actual benefits, in others you might just end

up paying too much money for marketing gimmicks.

For instance, NAS drives might be rated for 24/7 operation, but I've

had several Samsung drives over the past eight years or so and most of

them were used in 24/7 operation with no ill effects. However, the

actual workload I put on them was very low (basically just storing

movies and TV shows, primarily), so although they've been running

for a long time, they don't really need to do all that much work.

Also they were always in a well-controlled and stable environment

with regards to temperatures, humidity, vibrations and so on.

As another example, the WD Red NAS drives are rated to be operated

at up to 70 deg C, whereas usually HDDs are rated up to 55 deg C or

maybe occasionally 60 deg C. This should come in handy when you

operate them in a badly ventilated and tightly packed NAS case

I imagine. For instance, even WD's expensive enterprise drives

(RE) are only rated for 55 deg C.

(source: their pdf spec sheets on WDC's website).

http://www.wdc.com/wdproducts/library/SpecSheet/ENG/2879-800002.pdf,

http://www.wdc.com/wdproducts/library/SpecSheet/ENG/2879-771444.pdf

There might also be countermeasures to harden the drive against

vibrations (relevant if you pack lots of them into a tight space),

or better error-correcting capabilities, higher MTBF, higher workload

tolerance, lower URE rates etc. It all depends on the specific drive,

so if you really need to know I recommend digging into the spec sheets.

Depending on the specific drive the differences to a normal desktop

drive might be small or significant. For instance, I've read several

opinions stating that the WD Red is basically just a WD Green with a

different firmware, although personally I'm not sure how true that

really is.

EDIT:

Your posts' formatting always seems like some kind of poem

'Deathstars' made me laugh hard LOL

Like i said, it was the same for me with samsung drives. Switched to WD as well but they let me down sadly...

Hehe, happy to entertain. And yeah, I use manual line breaks,

it's a habit I've carried over from coding, although in an

ideal world I'd format it more evenly so that the line ends

line up almost perfectly (but who has the time for that ).

Edited May 7, 2014 by alpenwasser