Study: A Look At Hard Drive Reliability In Russia

[cmartin011]

i own a 500gb 7200.12 drive and its reliable owned it bout 2 years now when they first came out still quite and vibrationless. had tons of maxtor's die on me and others, i am running a samsung 1tb and its been chugging along with the seagate like no end lol

 

[Chewie]

I moved from Seagate to WD after WD replaced an external drive for me with minimal fuss.
However, this article does lend a slight emphasis to me getting some kind of HDD cooling sorted.

 

[ubercake]

I've tried Western Digital, Seagate, Maxtor (bought by Seagate) and Hitachi in my lifetime.

I've owned and built my own PCs from the early 1990s.

I've never had a Western Digital drive failure. One of my PCs is an eight year old P4 2.8 with the same Caviar SE OEM drives I put into it when I built it.

Before Seagate bought Maxtor, they were a definite competitor of Western Digital from a quality standpoint. Once Seagate incorporated Maxtor's production, they have become unreliable.

Every Maxtor I ever owned failed within a year. Maxtor was always the cheapest, so the sold high volumes. The quality was never there though.

I have a pre-Maxtor acquisition Seagate external USB drive from 2004 that is running fine to this day.

Every Hitachi I've owned failed within 3 years.

Considering the stats shown in the article, whenever you see a drive failure or hear of one, I'll bet it's hardly ever a Western Digital.

I recently bought an Intel SSD for my newest desktop. Hoping this will maintain a level of reliability comparable to their processors? I'm betting on the fact that there are no longer moving parts to contend with.

Western Digital is just getting into SSD. I look forward to getting some of their drives when they become more affordable. I'm hoping Western Digital continues to stress quality/reliability into the future.

 

[cookoy]

I shifted to WD drives as my Seagate drives were more prone to firmware failures from electrical spikes and power fluctuations. And i stay away from newer higher capacity drives. Just not risking putting too much eggs on one basket.

 

[bildo123]

I'm actually surprised my WD Raptor 10K (74GB) is still kicking(about 7 years near daily use). I ran the Speedfan SMART lookup and it came back decent. It's loud as hell though when accessing data...really time for a replacement before it goes.

 

[Proximon]

There are so many factors only lightly touched on here.
It's a data recovery company. Their clients are people who store important info on a HDD and then fail to back it up.
Geography and climate, operating conditions, power grid conditions and the other hardware associated with the drive are all going to be factors.
Still, I do find the description of failures interesting, and the differences between brands of the nature of those failures is actually useful.

 

[jojesa]

[citation][nom]superhighperf[/nom]this is totally opposite of my experience with HDD. i have had 4 Hitachi drives and they all failed. [/citation]
Same results here had two Hitachi laptop drives and one 1TB desktop drive failed. Have 6 Seagate drives (finger crossed)and had total of 14 and none failed. Also had 4 Western Digital and only one failed.

 

[Darkerson]

[citation][nom]thuan[/nom]try "seagate 7200.12 CC34 firmware reallocated sector count on" on google. I have two 500GB drives from Seagate that have the same problem. The one I have with my desktop at home I sold after a few hundred bad sector, the one I have at work I am still using and it has 4574 bad sectors, still going but very slow sometimes. It triggered Seatools of course, but I'm too busy to bring the damn thing to warranty without having another spare, so backing up everyday is now what I do. IMO 7200.12 with CC34 firmware is not that great either. FYII have another Seagate 7200.12 CC35 firmware and it has been ok, so it might be another firmware fiasco that Seagate doesn't want to acknowledge after the last generation "Seagate" case.[/citation]

Why would you sell someone a faulty hard drive? Thats just kinda of cruel and malicious. Anyway, I have the infamous 1.5 TB Seagate 7200.11. Luckily, mine had the updated firmware, but I have never really trusted it completely after everything that went down. The only reason I got it was because my friends kept telling me about how good Seagate was. Before, I always used Western Digital and always had good experiences. I guess when it comes time to get another one, I will be going back to WD, or i may give Hitachi a chance. Decisions decisions...

 

[sublifer]

Kind of surprised how well hitachi did. I'm not at all surprised how dismal seagate looked or how well WD did. I personally only buy WD drives but may have to start considering Hitachi now.

 

[sublifer]

I just went back and read through all the comments. It looks like a lot of people have had good experiences with WD like I have. One thing I found surprising is that there weren't any die-hard seagate fans calling this bunk or anything... 4 or so years ago there would have been. As other people pointed it, it really shows how far Seagate has fallen. Also wanted to add, IBM sold its hdd unit after they deskstar drives earned the nickname "deathstar" for extremely high failure rates. As a result I avoided hitachi's deskstar drives fearing they were the same quality. I'm glad it seems that fear was unfounded.

 

[cadder]

Years ago our supplier told us that he had the best luck with Fujitsu and IBM hard drives. (IBM now sold to Hitachi.) Looks like this still holds true. Shortly after that though Fujitsu dropped out of the home market so that leaves only Hitachi.

 

[thuan]

Why would you sell someone a faulty hard drive? Thats just kinda of cruel and malicious. Anyway, I have the infamous 1.5 TB Seagate 7200.11. Luckily, mine had the updated firmware, but I have never really trusted it completely after everything that went down. The only reason I got it was because my friends kept telling me about how good Seagate was. Before, I always used Western Digital and always had good experiences. I guess when it comes time to get another one, I will be going back to WD, or i may give Hitachi a chance. Decisions decisions...

Well it's not, because I sold it to a company who specialize in buying faulty HD, around 60% of the original price, so I didn't hesitate.

 

[duanes1967]

I wonder if the the prevalence of Seagate drives in servers didn't affect the failure statistics. I suspect that many failures are simply not reported. The fact that Seagate dominates in high capacity servers means that it is much more likely to be reported and that advanced data recovery is attempted.

Nobody cares if your crappy desktop Hitachi or Samsung drive fails. But the world stops when the server disks bite the big one!

 

[Guest]

The Hibachi 7k320

I just returned my Hitachi 7k320, it ran hot and died. God bless my data.

Suggest we call it "the hibachi".

Why would I trust anything that actually *needs* a three-year warranty? If a product needs the warranty, most products go in the first month. No hard drive lasts three years.

The Hibachi. lol

Kevin

 

[hardware specialists]

as a computer tech i know SEAGATES are garbage . western digital is where its at

 

[akorzan]

The 7200.9 was very reliable though... they messed up with the 7200.10 and 11 series.
Even going with Newegg reviews I remember that almost all the 9 series had 5 stars and the newer series had 3 or 4. You also cannot forget that the 9 series introduced perpendicular recording and they were very quiet compared to the competition.

Western Digital drives during the late 1990s and early 2000s were relatively louder compared to Seagate... but now they surpassed Seagate in that department too!

 

[Onus]

Proximon makes a valid point. Furthermore, even if some drives are more tolerant than others, a failure due to hidden abuse (like power problems) is hardly the fault of the drive.
I got a DOA maybe three years ago, but I can't remember the last failure in use where physical damage (e.g. a snapped-off SATA connector) was not involved. Having read all the comments and reviews, I never risked a 7200.11 drive, but 7200.10, 7200.12, WD Blue and WD Black have all been reliable. My wife now has the 150GB Velociraptor that I used for over a year (two now?) and it is still as quiet, fast, and cool-running as ever.

 

[Otus]

You average age analysis does not seem very robust. Did you consider the fact that a newer model (like Seagate 7200.11) cannot have as old drives as an older model (like 7200.10). Similarly, with regards to WD the price of over 500GB hard drives was much higher a few years ago, so most of them would have been purchased quite recently.

 

[Darkerson]

[citation][nom]thuan[/nom]Well it's not, because I sold it to a company who specialize in buying faulty HD, around 60% of the original price, so I didn't hesitate.[/citation]

It just came across as you selling it to some random person or something. I didn't know places like that existed. In that case, cant blame you one bit. Ill have to keep that in mind if my Seagate ever craps out on me and they try and get out of replacing it or whatnot.

 

[lamorpa]

Please send your amusing anecdotes to Reader's Digest. Anyone's personal experiences with a couple of hard drives are absolutely statistically insignificant. Read this article and the studies by EMC and Google.

 

[pinkfloydminnesota]

It looks like you're using the term "failure rate" when you mean "percentage of all failing hard drives".

 

[g00ey]

I really don't like the praise here over the WD hard drives, they are really unreliable and have very short service life. Sure, in the past they were pretty decent but not anymore. Look at this article:

http://www.pcnsdfw.com/bad-drives.html

I have also read on forums several people who have lost their drives and WD is a common drive to fail in these forums. I'm really doubtful about this study. Modern hard drives rarely crash when they die, they usually get the click-of-death, then it is kiss goodbye to that drive and a cost of a few thousand dollars to recover the data.

 

[TeraMedia]

I have had 4 of 4 WD 400 GB RE drives fail, all after the 3-year warranty.
I have 2 WD 5000YS drives (500 GB RE or RE2, not sure which) that are beginning to flake after 4+ years. I have 2 WD 5000ABYS drives (500 GB RE2) that are doing fine after 3 years. And I have 1 WD 500 GB RE3 (not ABYS, but not sure of the model name) that I bought recently to replace one of the 5000YS drives, and that is working fine.

All of these have resided (until failure) in an HTPC / video server (warm inside) running 24x7.

In my opinion, the 400 GB drives lasted long enough to be considered obsolete. By the time they failed, I could have replaced them all with two 1.5 TB drives running in RAID 1 instead of 4 in RAID 5, and the performance and reliability would have been similar, but with easier and faster RAID recovery, more storage, less heat, fewer parts, lower cost and less space in the case.

 

[insightdriver]

I have always been drive-agnostic. The reason is, I know that hard drives will fail. Since I know that and back up my data regularly, then how reliable a brand doesn't really matter for me, a home user. My HDD's are not uses 24/7 as they would be in a data center. I just see a natural inclination for people to have brand loyalty. I guess sitting in front of a tv as a child, bombarded with advertising has made good little consumers with brand loyalty.

 

[g00ey]

[citation][nom]TeraMedia[/nom]... drives running in RAID 1 instead of 4 in RAID 5, and the performance and reliability would have been similar, but with easier and faster RAID recovery, more storage, less heat, fewer parts, lower cost and less space in the case.[/citation]
Hardware RAID is really unreliable, it's a Sysadmins nightmare. The problem is that the errors are handled internally in the controller without you/the system knowing it. Once the errors starts to get visible it usually is too late to do anything about it. Most Sysadmins say stay away from RAID5.

Also it is better to use double redundancy; a very common scenario is when one drive fails and is replaced another drive fails during the rebuild of the storage pool due to the higher load that is put onto the drives during the rebuild process. Therefore it is recommended to use at least double redundancy.

I truly recommend using ZFS in raidz2 mode which compares to RAID6 instead which is much more reliable and easy to administrate.

RAID1 is more reliable than RAID5 but it is a lot better to use a mirrored storage pool in ZFS instead of RAID1.

ZFS also have better expansion capabilities than any hardware RAID can get. You can also tweak the IOPs of a ZFS storage pool by adding a couple of SSDs for read and write cache.

I don't want to sound like a salesman, which I'm not but ZFS is the most reliable filesystem available today. Stock exchanges and CERN use ZFS for this single reason. btrfs is another filesystem that is thought to be an alternative to ZFS but it is many years behind in terms of development. Linux has had problems to properly implement ZFS due to incompatibility issues with the CDDL licensing. ZFS is available for Solaris, OpenSolaris, FreeBSD and FreeNAS (which is a FreeBSD distro).