Stories
Slash Boxes
Comments

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

Disk Drive Failures 15 Times What Vendors Say

Posted by Zonk on Fri Mar 02, 2007 04:15 PM
from the cough-sputter-wheeze-choke dept.
jcatcw writes "A Carnegie Mellon University study indicates that customers are replacing disk drives more frequently than vendor estimates of mean time to failure (MTTF) would require.. The study examined large production systems, including high-performance computing sites and Internet services sites running SCSI, FC and SATA drives. The data sheets for the drives indicated MTTF between 1 and 1.5 million hours. That should mean annual failure rates of 0.88%, annual replacement rates were between 2% and 4%. The study also shows no evidence that Fibre Channel drives are any more reliable than SATA drives."
+ -
story
This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More
Loading... please wait.
  • Repeat? (Score:2, Insightful)

    Didn't we already see this evidence with Google's report?
    • Re:Repeat? (Score:4, Informative)

      by georgewilliamherbert (211790) on Friday March 02 2007, @04:20PM (#18211790)
      We did both this study and the Google study in the first couple of days after FAST was over. Completely redundant....
    • Re: (Score:2, Interesting)

      Yes, and its mentioned in the report.
      The best part about the entire thing is the very last quote:

      "If they told me it was 100,000 hours, I'd still protect it the same way. If they told me if was 5 million hours I'd still protect it the same way. I have to assume every drive could fail."

      Just common sense.
      • Re:Repeat? (Score:5, Informative)

        by ajs (35943) <ajs@NosPAm.ajs.com> on Friday March 02 2007, @04:34PM (#18211992) Homepage Journal

        The best part about the entire thing is the very last quote:

        "If they told me it was 100,000 hours, I'd still protect it the same way. If they told me if was 5 million hours I'd still protect it the same way. I have to assume every drive could fail."

        Just common sense.
        It's "common sense," but not as useful as one might hope. What MTTF tells you is, within some expected margin of error, how much failure you should plan on in a statistically significant farm. So, for example, I know of an installation that has thousands of disks used for everything from root disks on relatively drop-in-replaceable compute servers to storage arrays. On the budgetary side, that installation wants to know how much replacement cost to expect per annum. On the admin side, that installation wants to be prepared with an appropriate number of redundant systems, and wants to be able to assert a failure probability for key systems. That is, if you have a raid array with 5 disks and one spare, then you want to know the probability that three disks will fail on it in the, let's say, 6 hour worst-case window before you can replace any of them. That probability is non-zero, and must be accounted for in your computation of anticipated downtime, along with every other unlikely, but possible event that you can account for.

        When a vendor tells you to expect 1 0.2% failure rate, but it's really 2-4% that's a HUGE shift in the impact to your organization.

        When you just have one or a handful of disks in your server at home, that's a very different situation from a datacenter full of systems with all kinds of disk needs.
        • just assume 3 years (Score:5, Informative)

          by crabpeople (720852) on Friday March 02 2007, @05:05PM (#18212416) Journal
          A good rule of thumb is 3 years. Most hard drives fail in 3 years. I dont know why, but im currently seeing alot of bad 2004 branded drives and consider that right on schedule. Last year the 02-03 drives were the ones failing left and right. I just pulled one this morning thats stamped march 04. Just started acting up a few days ago. Like clockwork.

        • Re:Repeat? (Score:4, Informative)

          by ShakaUVM (157947) on Friday March 02 2007, @08:30PM (#18214254) Homepage Journal
          Except MTBF is just pulled out of their asses. Look at the development cycle of a hard drive. Look at the MTBF. I used to work for an engineering company, and have worked doing test suites to determine MTBF. Sure, there's numbers involved, but it's probably 60% wishful thinking and 40% science.

          Believe me, they aren't determining an 11 year MTBF empirically.
          • There's also a non-zero probability that all of the air molecules in a room will rush to the corner of the room, suffocating the occupants.
            • Just think what a fantastic way to die that would be. You'd get all kinds of notoriety.
  • by User 956 (568564) on Friday March 02 2007, @04:17PM (#18211744) Homepage
    The data sheets for the drives indicated MTTF between 1 and 1.5 million hours.

    Yeah, but I bet they didn't say what planet those hours are on.
    • Or what percentage of the speed of light they were traveling.
    • If an observer on a rail platform measures the MTF of a hard disk on a rail car moving at speeds close to the speed of light...

    • Yes, I am SHOCKED that companies have implemented a systematic program of distorting the truth in order to increase profits.

      I propose a new term for the heinous practice---"marketing".
      • by Beardo the Bearded (321478) on Friday March 02 2007, @04:43PM (#18212126)
        What, really?

        The same companies that lie about the capacity on EVERY SINGLE DRIVE they make? You don't think that they're a bunch of lying fucking weasels? (We're both using sarcasm here.)

        I don't care how you spin it. 1024 is the multiple. NOT 1000!

        Failure doesn't get fixed because making a drive more reliable means it costs more. If it costs more, it's not going to get purchased.

        • Re: (Score:3, Informative)

          Before computers were used in real engineering, we could get away with "k" sometimes meaning 1024 (like in memory addresses) and sometimes meaning 1000 (like in network speeds). Those days are past. Now that computers are part of real engineering work, even the slightest amount of ambiguity is not acceptable .

          Differentiating between "k" (=1000) and "ki" (=1024) is a sign that the computer industry is finally maturing. It's called progress.

          • I'd just like to point out that computers were used for "real" engineering long before they became ubiquitous in the workplace or home. Why do you think FORTRAN is one of the oldest computing languages in existence?
        • Re: (Score:3, Informative)

          Standard metric is indeed powers of 10, and a megabyte is indeed 10^6 bytes.

          To clear up the confusion, the notation for binary, as in 2^20 bytes was developed. That would be a Mebibyte.

          http://en.wikipedia.org/wiki/Mebibyte [wikipedia.org]

        • by binarybum (468664) on Friday March 02 2007, @06:46PM (#18213536) Homepage
          yeah, I used to think they were dirty bastards, but they just work on a different scale than the rest of us.
              The trick is to purchase your HD in pennies.

            "100,000 pennies! why that's 1024 dollars!!"
  • In other news... (Score:5, Informative)

    by Mr. Underbridge (666784) on Friday March 02 2007, @04:22PM (#18211808)
    ...Carnegie Mellon researchers can't tell a mean from a median. This is inherently a long-tailed distribution in which the mean will be much higher than the median. Imagine a simple situation in which failure rates are 50%/yr, but those that last beyond a year last a long time. Mean time to failure might be 1000 years. You simply can't compare the statistics the way they have without knowing a lot more about the distribution than I saw in the article. Perhaps I missed it while skimming.
  • I believe it... (Score:3, Informative)

    by madhatter256 (443326) on Friday March 02 2007, @04:23PM (#18211832)
    Yeh. Don't rely on the HDD after it surpasses its' manufacturer warranty.
    • Re: (Score:2, Insightful)

      Also, don't rely on the HDD before it surpasses its manufacturer warranty. All the warranty means is you get a replacement if it breaks - it doesn't provide any extra guarantees of the disk not failing.
  • Disk Drive Failures 15 Times What Vendors Say [...] That should mean annual failure rates of 0.88% [but] annual replacement rates were between 2% and 4%.

    0.88 * 15 = 4?
  • by Lendrick (314723) on Friday March 02 2007, @04:26PM (#18211880) Homepage Journal
    In the article, they mention that the study didn't track actual failures, just the how often customers *thought* there was a failure and replaced their drive. There are all sorts of reasons someone might think a drive has failed. They're not all correct. I can't begin to guess what percentage of those perceived failures were for real.

    This study is not news. All it says is that people *think* their hard drives fail more often than the mean time to failure.
    • And I think they fail less often than the MTTF. There, the statistics are satisfied as well, and it's still not news.
    • by crabpeople (720852) on Friday March 02 2007, @05:12PM (#18212520) Journal
      Thats fair, but if you pull a bad drive, ghost it (assuming its not THAT bad), plop the new drive in, and the system works flawlessly, what are you to assume?

      I dont really care to know exactly what is wrong with the drive. If i replace it, and the problem goes away, I would consdier that a bad drive. Even if you could still read and write to it. I just did one this morning that showed no symptoms other than windows taking what I considered a long time, to boot. All the user complained about was sluggish performance, and there were no errors or drive noises to speak of. Problem fixed, user happy, drive bad.

      As I already posted, a good rule of thumb is 3 years from the date of manufacture, is when most drives go bad.

  • by neiko (846668) on Friday March 02 2007, @04:30PM (#18211936)
    TFA seems surprised by SATA drives lasting as long as Fibre...why one earth would your data interface have any consequences on the drive internals? Or are we talking assuming Interface = Data Throughput?
    • Re: (Score:3, Insightful)

      TFA seems surprised by SATA drives lasting as long as Fibre...why one earth would your data interface have any consequences on the drive internals? Or are we talking assuming Interface = Data Throughput?

      That statement is based on the long-held assumption that hard drive manufacturers put better materials and engineering into enterprise-targeted drives [Fibre] than they put into consumer-level drives [SATA].

      Guess not...

      • Re: (Score:3, Informative)

        They certainly charge enough more. SATA drives run about $0.50 per gig. Comparable Fibre Channel drives run about $3 per gig. A sensible person would expect the Fibre Channel drive to be as much as 6 times as reliable, but per the article there is no difference.
    • by mollymoo (202721) on Friday March 02 2007, @04:42PM (#18212110) Journal

      TFA seems surprised by SATA drives lasting as long as Fibre...why one earth would your data interface have any consequences on the drive internals?

      Fibre Channel drives, like SCSI drives, are assumed to be "enterprise" drives and therefore better built than "consumer" SATA and PATA drives. It's nothing inherent to the interface, but a consequence of the environment in which that interface is expected to be used. At least, that's the idea.

  • by Danga (307709) on Friday March 02 2007, @04:30PM (#18211940)
    I have had 3 personal use hard drives go bad in the last 5 years, they were either Maxtor or Wester Digital. I am not hard on the drives other than leaving them on 24/7. The drives that failed were all just for data backup and I put them in big, well ventilated boxes. With this use I would think the drives would last for years (at least 5 years), but nope! The drives did not arrive broken either, they all functioned great for 1-2 years before dying. The quality of consumer hard drives nowadays is way, WAY low, and the manufacturers should do something about it.

    I don't consider myself a fluke because I know quite a few other people who have had similar problems. What's the deal?

    Also, does anyone else find this quote interesting?:

    "and may have failed for any reason, such as a harsh environment at the customer site and intensive, random read/write operations that cause premature wear to the mechanical components in the drive."

    It's a f$#*ing hard drive! Jesus H Tapdancing Christ how can they call that premature wear, do they calculate the MTTF by just letting the drive sit idle and never reading and writing to it? That actually wouldn't suprise me.
        • Re: (Score:3, Insightful)

          The biggest reason I think it would make no difference is because if unconditioned power is supposed to be so bad for electronics then why is the only thing that I have a problem with turn out to be the hard drives? I would think bad power would take out RAM before a hard disk.

          Ram has no significant inductive load.
  • Check SMART Info (Score:4, Interesting)

    by Bill Dimm (463823) on Friday March 02 2007, @04:41PM (#18212094) Homepage
    Slightly off-topic, but if you haven't checked the Self-Monitoring, Analysis and Reporting Technology (SMART) info provided by your drive to see if it is having errors, you probably should. You can download smartmontools [sourceforge.net], which works on Linux/Unix and Windows. Your Linux distro may have it included, but may not have the daemon running to automatically monitor the drive (smartd).

    To view the SMART info for drive /dev/sda do:
    smartctl -a /dev/sda
    To do a full disk read check (can take hours) do:
    smartctl -t long /dev/sda

    Sadly, I just found read errors on a 375-hour-old drive (manufacturer's software claimed that repair succeeded). Fortunately, they were on the Windows partition :-)
    • Re: (Score:3, Informative)

      Slightly off-topic, but if you haven't checked the Google paper on Self-Monitoring, Analysis and Reporting Technology (SMART) info provided by your drive to see if it is having errors, you probably should. The paper is available here: http://hardware.slashdot.org/hardware/07/02/18/04 2 0247.shtml [slashdot.org]

      The conclusions are roughly the following: a) if there are SMART errors, the disk will fail soon, b) if there are no SMART errors, the disk is still likely to fail. They saw no SMART errors on 36% of their failed d
  • New meaning for RAID: Redundant Articles of Identical Discourse.
    Slashdot has a high rate of RAID, which is a bad thing. Which is a bad thing. It has been a whole 9 days. Slashdot needs a story moderation system so dupe articles can get modded out of existance. Ditto for slashdot editors who do the duping! :) (I have long since disabled tagging since 99% of the tags were completely worthless: "yes", "no", "maybe", "fud", etc. If tagging is actually useful now, please let me know!)

    Can we get redundant posting on the story about google's paper [slashdot.org]?
  • Seagate (Score:4, Insightful)

    by mabu (178417) on Friday March 02 2007, @05:04PM (#18212390)
    After 12 years of running Internet servers, I won't put anything but Seagate SCSI drives in any mission critical servers. My experience indicates Seagate drives are superior. Who's the worst? Quantum. The only thing Quantum drives are good for is starting a fire IMO.
  • by dangitman (862676) on Friday March 02 2007, @05:09PM (#18212478)
    Pick any two.

    I've noticed this personally. Now, anecdotal evidence doesn't count for a lot, and it may be a case that we are pushing our drives more. But back in the day of 40MB hard drives that cost a fortune, they used to last forever. The only drive I ever had fail on me in the old days were the Syquest removable HD cartridges, for obvious reasons. But even they didn't fail that often, considering the extra wear-and-tear of having a removable platter with separate heads in the drive.

    But these days, with our high-capacity ATA drives, I see hard drives failing every month. Sure, the drives are cheap and huge, but they don't seem to make them like they used to. I guess it's just a consequence of pushing the storage and speed to such high levels, and cheap mass-production. Although the drives are cheap, if somebody doesn't back up their data, the costs are incalculable if the data is valuable.

  • by Tim Browse (9263) on Friday March 02 2007, @05:38PM (#18212852)

    ...is that it detects SMART disk errors in normal use (i.e. you don't have to be watching the BIOS screens when your PC boots).

    When I was trying the Vista RC, it told me that my drive was close to failing. I, of course, didn't believe it at first, but I ran the Seagate test floppy and it agreed. So I sent it back to Seagate for a free replacement.

    About the only feature that impressed me in Vista, sadly. (And I'm not sure it should have impressed me, tbh. I'm assuming XP never did this as I've never seen/heard of such a feature.)

    • by Matt Perry (793115) on Friday March 02 2007, @09:03PM (#18214434)

      When I was trying the Vista RC, it told me that my drive was close to failing. ... About the only feature that impressed me in Vista, sadly.
      Be sad no more. SmartMonTools [sourceforge.net] will run in UNIX or Windows and notify you if it detects SMART errors. For the Windows installer look for the phrase "Install the Windows package" on the smartmontools home page..
  • by CorporalKlinger (871715) on Friday March 02 2007, @09:08PM (#18214468)
    I think one of the key problems here isn't necessarily the statistical methods used, it is that the CMU team was comparing real-life drive performance to the "ideal" performance levels predicted by the drive manufacturers. Allow me to provide two examples of this "apples to oranges" comparison problem.

    I have had two computers with power supply units that were "acting up." They ended up killing my hard drives on multiple occasions - Seagates, WD's, Maxtors, etc. It didn't matter what type of drive you put in these systems, the drive would die after anywhere from a week to two years. I later discovered that the power supplies were the problems, replaced them with brand new ones, and replaced the drives one last time. That was quite some time ago (years), and those drives, although small, still work, and have been transferred into newer computer systems since that time. The PSU was killing the drives; they weren't inherently bad or had a manufacturing defect. A friend of mine who lives in an apartment building constructed circa 1930 experienced similar problems with his drives. After just a few months, it seemed like his drives would spontaneously fail. When I tested his grounding plug, I found that it was carrying a voltage of about 30V (a hot ground - how wonderful). Since he moved out of that building and replaced his computer's PSU, no drive failures.

    The same type of thing is true in automobile mileage testing. Car manufacturers must subject their cars to tests based on rules and procedures dictated by state and federal government agencies. These tests are almost never real world - driving on hilly terrain, through winds, with the headlights and window wipers on, plus the AC for defrost. They're based on a certain protocol developed in a laboratory to level the playing field and ensure that the ratings, for the most part, are similar. It simply means when you buy a new car, you can expect that under ideal conditions and at the beginning of the vehicle's life, it should BE ABLE to get the gas mileage listed on the window (based on an average sampling of the performance of many vehicles).

    My point is that there really isn't a decent way to go about ensuring that an estimated statistic is valid for individual situations. By modifying the environmental conditions, the "rules of the game" change. A data-center with exceptional environmental control and voltage regulation systems, and top-quality server components (PSU's, voltage regulators, etc.) should expect to experience fewer drive failures per year than the drives found in an old chicken-shack data center set up in some hillbilly's back yard out in the middle of nowhere where quality is the last thing on the IT team's mind. It's impractical to expect that EVERY data center will be ideal - and since it's very very difficult to have better than the "ideal" testing conditions used in the MTTF tests - the real-life performance can only move towards more frequent and early failures. Using the car example above, since almost nobody is going to be using their vehicle in conditions BETTER than the ideal dictated by the protocols set forth by the government, and almost EVERYONE will be using their vehicles under worse conditions, the population average and median have nowhere to go but down. That doesn't mean the number is wrong, it just means that it's what the vehicle is capable of - but almost never demonstrates in terms of its performance - since ideal conditions in the real world are SO rare.
    • Off-Topic: SI Units (Score:5, Informative)

      by ewhac (5844) on Friday March 02 2007, @05:21PM (#18212634) Homepage Journal

      I just can't believe that the same vendors that would misrepresent the capacity of their disk by redefining a Gigabyte as 1,000,000,000 bytes instead of 1,073,741,824 bytes would misrepresent their MTBF too!

      Not that this is actually relevant or anything, but there's been a long-standing schism between the computing community and the scientific community concerning the meaning of the SI prefixes Kilo, Mega, and Giga. Until computers showed up, Kilo, Mega, and Giga referred exclusively to multipliers of exactly 1,000, 1,000,000, and 1,000,000,000, respectively. Then, when computers showed up and people had to start speaking of large storage sizes, the computing guys overloaded the prefixes to mean powers of two which were "close enough." Thus, when one speaks of computer storage, Kilo, Mega, and Giga refer to 2**10, 2**20, and 2**30 bytes, respectively. Kilo, Mega, and Giga, when used in this way, are properly slang, but they've gained traction in the mainstream, causing confusion among members of differing disciplines.

      As such, there has been a decree [nist.gov] to give the powers of two their own SI prefix names. The following have been established:

      • 2**10: Kibi (abbreviated Ki)
      • 2**20: Mebi (Mi)
      • 2**30: Gibi (Gi)

      These new prefixes are gaining traction in some circles. If you have a recent release of Linux handy, type /sbin/ifconfig and look at the RX and TX byte counts. It uses the new prefixes.

      Schwab

    • Re: (Score:3, Interesting)

      I had a 4mb 72-pin parity SIMM go bad one time...this was about 12 years ago in a 486 I used to have. It just didn't work one day (it worked for the first two months). Turn the computer on, get past BIOS start, bam...parity error before bootloader could even start. Reboot, try again, parity error. Turn off parity checking, it actually started to boot and then crashed. The RAM was obviously very defective...when I took that 1 stick out the computer booted normally even with parity on, if I tried to boot
      • by Fulcrum of Evil (560260) on Friday March 02 2007, @06:32PM (#18213398)

        Well, the hard-drive makers are correct on the size thing - a Gigabyte is 1000 Megabytes, and the OS and software makers are wrong.

        Yeah, they coined the term and have been using it for 40 years, but they're wrong.

        Gigabytes are actually displayed as Gigabytes, or that the listing is changed to correctly display Gibibytes as the value? (or Kibibytes, Mebibytes, whatever)

        Listen, just because someone comes up with a standard doesn't obligate everyone to use it, especially when they already have a perfectly workable system already. Claiming that NIST can impose an unwanted standard on the world is like saying that it isn't a word until the OED lists it.