Backblaze posts failure rates

[Phynaz]

Don't buy Seagate - some models have failure rates over 25%.

http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/

 

[KingFatty]

Wow Hitachi looking pretty good.

If the price were right, we would be buying nothing but Hitachi drives. They have been rock solid, and have had a remarkably low failure rate.

 

[Cerb]

Surprising. I figured there'd be differences, such as the WD v. HGST (do current Toshibas follow that trend?), but not so much less from Seagate. I wonder how much is outright quality, as compared to intended environment?

 

[Phynaz]

2/3 of my recent drive purchases have been Hitachi, but they are just so darn slow.

On the other hand they are dead silent.

 

[ethebubbeth]

I'm so glad that BackBlaze publishes these kinds of statistics, and I wish other large datacenter companies would as well.

I feel vindicated that the Hitachi 3TB (HDS5C3030ALA630) drives that I have been using for years have the lowest failure rates, and that the the WD 3TB Red (WD30EFRX) drives I've purchased recently, despite have 3x the failure rate of the Hitachis, are still some of the better ones out there.

 

[Elixer]

Look at the sample size...

Seagate 12,765 39,576 1.4
Hitachi 12,956 36,078 2.0
Western Digital 2,838 2,581 2.5
Toshiba 58 174 0.7
Samsung 18 18 3.7

That means, between seagate & hitachi, hitachi is the clear winner.
WD's small sample size isn't showing that good either...
The 'popular' reds:

Western Digital Red (WD30EFRX) 3.0TB 346 0.5 3.2%

Isn't very good.
Too bad we don't see any data for blacks, or the RE edition.

 

[Ayah]

Look at the sample size...

That means, between seagate & hitachi, hitachi is the clear winner.
WD's small sample size isn't showing that good either...
The 'popular' reds:

Isn't very good.
Too bad we don't see any data for blacks, or the RE edition.

you wouldn't. they're ridiculously expensive compared to "consumer" drives, and don't fit backblaze's storage paradigm.

 

[Soulkeeper]

I guess this means my ST3000DM001 has a 9.8% annual failure rate ....

good thing I have 2 of them (each with different country of origin/firmware) and do manual backups

 

[Charlie98]

2/3 of my recent drive purchases have been Hitachi, but they are just so darn slow.

On the other hand they are dead silent.

I've got 4 Hitachi HDDs, I like them a lot... but they are slow. I've got one in an external enclosure... it's not very quiet, either.

My original 2TB Red went belly up, WD replaced it with a 3TB Red refurb... hopefully this one will last...

 

[PliotronX]

Glad to see it in black and white, Seagate suxx0rs.

 

[lehtv]

You can find the model numbers used in the study here: http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/

The drive with the highest fail rate, ST31500341AS (25.4%), is a Seagate 7200.11 drive from 2008. A six years old model using three 500GB platters. This makes me doubt there are any useful conclusions to be drawn from this drive's survivability when compared to newer 1TB per platter drives. It is clearly an outlier which you can't use to judge Seagate's reliability in general, especially that of their current drives. The second highest failure rate was on ST31500541AS (9.8%), much improved over the -341AS. But even this one is a five year old model, not exactly relevant any more. For example, ST4000DM000 is a new drive with 3.8% failure rate, similar to WD's numbers.

I don't think these results should be taken as "Seagate sucks, Hitachi rocks", one has to look at each model separately instead of blindly trusting one brand over another, and also weigh the cost of the drive and the length of its warranty against the likelihood of it failing.

Also, importantly, the study did not include many of the popular general purpose home PC drives, sorted here in order of number of reviews on newegg:

1. WD Black WD1002FAEX
2. WD Blue WD10EZEX
3-6. Smaller WD Blue drives
7. (was included) ST3000DM001
8. ST1000DM003
9. ST2000DM001
10. WD Black WD2002FAEX

Instead, mostly lower RPM drives or older 7200RPM drives were in, and results for either of those may not apply to new 7200 RPM drives. I would also love to see results for Toshiba hard drives.

 

[VirtualLarry]

You can find the model numbers used in the study here: http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/

The drive with the highest fail rate, ST31500341AS (25.4%), is a Seagate 7200.11 drive from 2008. A six years old model using three 500GB platters.
...
I would also love to see results for Toshiba hard drives.

That particular Seagate model was the one with all of the firmware-update troubles, was it not? If that's the one, then no wonder it's the poster child for unreliability.

I would have loved to see more on the Toshiba 3TB 7200 RPM drives too.

 

[Phynaz]

Also, importantly, the study did not include many of the popular general purpose home PC drives, sorted here in order of number of reviews on newegg:

1. WD Black WD1002FAEX
2. WD Blue WD10EZEX
3-6. Smaller WD Blue drives
7. (was included) ST3000DM001
8. ST1000DM003
9. ST2000DM001
10. WD Black WD2002FAEX

Instead, mostly lower RPM drives or older 7200RPM drives were in, and results for either of those may not apply to new 7200 RPM drives. I would also love to see results for Toshiba hard drives.

Blacks are hardly popular general home use drives.

 

[lehtv]

Blacks are hardly popular general home use drives.

Why would you say that? I would say Blacks are very often bought for home use by people who think they are more reliable than Blues or just like to have longer warranty. Plenty of gamers with Black drives too.

 

[ashetos]

Yeah, blacks are popular due to the longer warranty. This study explains why Seagate drive have such short drive warranties.

 

[PliotronX]

It's not just models or sampling size or a generalized failure rate. When I service a system with a failing hard drive, it's usually a Seagate and when they fail it has usually been catastrophic and recovering data complex. By contrast when a WD fails it has usually worked enough to be able to get data off before it just doesn't work. This has been IME at least, I hear of some people having the reverse experience but you won't find Seagate in my personal systems.

 

[evilspoons]

The drive with the highest fail rate, ST31500341AS (25.4%), is a Seagate 7200.11 drive from 2008. A six years old model using three 500GB platters. This makes me doubt there are any useful conclusions to be drawn from this drive's survivability when compared to newer 1TB per platter drives. It is clearly an outlier which you can't use to judge Seagate's reliability in general, especially that of their current drives. The second highest failure rate was on ST31500541AS (9.8%), much improved over the -341AS. But even this one is a five year old model, not exactly relevant any more. For example, ST4000DM000 is a new drive with 3.8% failure rate, similar to WD's numbers.

I don't think these results should be taken as "Seagate sucks, Hitachi rocks", one has to look at each model separately instead of blindly trusting one brand over another, and also weigh the cost of the drive and the length of its warranty against the likelihood of it failing.

Ahh, a voice of reason. Thank you!

It's nice to see my two new ST4000DM000 drives aren't very different from the WD units I could have bought instead rather than worrying about Seagate's overall higher failure rate.

From the very same article:

We are focusing on 4TB drives for new pods. For these, our current favorite is the Seagate Desktop HDD.15 (ST4000DM000).

 

[BrightCandle]

What I would like to see them do is take the failure rate with age graph and split it down to individual drives. From the average data for drives it looks like they have had a couple of very bad Samsung drives but that isn't necessarily the case for all the models. I would certainly like to see a comparison involving drive model over time as this is clearly having a big impact on the data right now.

 

[Phynaz]

Why would you say that? I would say Blacks are very often bought for home use by people who think they are more reliable than Blues or just like to have longer warranty. Plenty of gamers with Black drives too.

General home use means Dell / HP etc. They aren't using blacks.

 

[lehtv]

General home use means Dell / HP etc. They aren't using blacks.

Obviously I was referring to custom builds which is the only type of build you see reflected in Newegg user reviews, and also the only type of build that matters when considering which hard drive to buy.

Also you can add a Black drive into an OEM build yourself.

 

[lehtv]

Yeah, blacks are popular due to the longer warranty. This study explains why Seagate drive have such short drive warranties.

No, it doesn't. Seagate drives typically have the same 2 year warranty as WD Blue. Toshiba has 1-2 year warranty depending on the drive. And WD Black's 5 year warranty doesn't mean it's more reliable, it just makes it more expensive.

 

[pauldun170]

No, it doesn't. Seagate drives typically have the same 2 year warranty as WD Blue. Toshiba has 1-2 year warranty depending on the drive. And WD Black's 5 year warranty doesn't mean it's more reliable, it just makes it more expensive.

My experience with WD green and blue, also considering user reports is that they crap out around the 2 year mark with heavy usage.

WD Blacks are warrantied for 5 years. My experience with WD blacks is that every single one that I have ever purchased is still running today. Several are over 5 years old and have been used for multiple OS installs, file servers, app servers all sorts of crap.

Initial cost
WD Black 2TB - $150
WD Blue 2TB - $110
WD Green 2TB - $90

2TB cost over 5 years based on assumption drives will fail
WD Black - $150
WD Blue - $275
WD Green - $225

In my view, WD Black drives are the cheaper option

 

[evilspoons]

WD Black's 5 year warranty doesn't mean it's more reliable, it just makes it more expensive

The warranty means the manufacturer trusts the device to not fail within that period, otherwise they would be losing a huge pile of money. It's a reasonable indicator of a manufacturer's trust in their product offerings.

My main problem with WD Black drives is they're much louder than WD Green. I noticed a significant amount more audible chatter on my storage drive when a WD Green failed and I replaced it with a WD Black.

 

[ashetos]

No, it doesn't. Seagate drives typically have the same 2 year warranty as WD Blue. Toshiba has 1-2 year warranty depending on the drive. And WD Black's 5 year warranty doesn't mean it's more reliable, it just makes it more expensive.

Most Seagate drives in my country have 1 year warranty, and I can't find warranty information in Seagate site easily (I wonder why).

If you use WD Black and Red drives with 3 and 5 year warranties, you can actually save some money, considering Seagate drives seem to fail around the 2 year mark (we don't buy seagate in my organization anymore).

For fairness, I have to say that WD Green drives have pretty bad reliability.

 

[lehtv]

The warranty means the manufacturer trusts the device to not fail within that period, otherwise they would be losing a huge pile of money. It's a reasonable indicator of a manufacturer's trust in their product offerings.

Why would they be losing money if the product costs more to begin with? It's a trade-off. The cost of the Black drive is higher because it is covered by a longer warranty. Just like it would cost you extra to buy an extended warranty for a laptop, in this case the extended warranty is just already part of the product.

Warranty alone doesn't say much about reliability. Of course, you can say there's a general trend that longer warranty correlates with higher reliability, but you can't look at any particular product and say it is more reliable than another product based on the fact that it has longer warranty. For example, EVGA offers power supplies with 10 year warranty but they are not one bit more reliable than comparably powerful and efficient Seasonic units with 5 year warranty, in fact they may even be less reliable given that they cost less.