Slashdot Mirror

← Back to Stories (view on slashdot.org)

BackBlaze's Hard Drive Stats for Q2 2017 (backblaze.com)

Posted by msmash on from the reality-check dept.

BackBlaze is back with its new hard drive reliability report: Since our last report for Q1 2017, we have added 635 additional hard drives to bring us to the 83,151 drives we'll focus on. We'll begin our review by looking at the statistics for the period of April 1, 2017 through June 30, 2017 (Q2 2017). [...] When looking at the quarterly numbers, remember to look for those drives with at least 50,000 drive hours for the quarter. That works out to about 550 drives running the entire quarter. That's a good sample size. If the sample size is below that, the failure rates can be skewed based on a small change in the number of drive failures.

Editor's note: In short: hard drives from HGST, a subsidiary of Western Digital, and Toshiba were far more reliable than those from Seagate across the models BackBlaze uses in its datacenters.

68 of 99 comments (clear)

  1. HGST and Toshiba have been at the top for years by CrashNBrn · · Score: 4, Informative

    Editor's note: In short: hard drives from HGST, a subsidiary of Western Digital, and Toshiba were far more reliable than those from Seagate

    Which has been the case since BackBlaze started releasing it's reliability numbers, aside from a few instances where a specific model of Seagate performed unusually well.

    1. Re:HGST and Toshiba have been at the top for years by Thelasko · · Score: 2

      HGST has come a long way since the Deathstar days.

      --
      One of our competitors trademarked the term "hypothesis". From now on, we will call them "boneheaded ideas".
    2. Re:HGST and Toshiba have been at the top for years by fred6666 · · Score: 2

      and yet, they continue buying far more Seagate than anything else.
      The reason is that they are cheaper, and it's not worth it to pay a premium for other brands.

      On amazon.ca, 4TB Seagate is $140, Western Digital is $157.

      For external (don't understand why, but it's cheaper) 8TB, Seagate is $235 and Western is $280

    3. Re:HGST and Toshiba have been at the top for years by thegarbz · · Score: 1

      HGST has come a long way since the Deathstar days.

      From your own link HGST never manufactured a "Deathstar". That was entirely IBM's doing.

    4. Re:HGST and Toshiba have been at the top for years by Thelasko · · Score: 2

      From your own link HGST never manufactured a "Deathstar". That was entirely IBM's doing.

      "The line was continued by Hitachi when in 2003 it bought IBM's hard disk drive division and renamed it Hitachi Global Storage Technologies. In 2012 Hitachi sold the division to Western Digital who rebranded it as HGST."

      Hitachi didn't buy just the IP, they bought the division. That typically means all of the IP, manufacturing facilities, engineering, and business operations related to a product line. It's likely that there are a few people that worked on the "deathstar" that still work for HGST today.

      --
      One of our competitors trademarked the term "hypothesis". From now on, we will call them "boneheaded ideas".
    5. Re:HGST and Toshiba have been at the top for years by fred6666 · · Score: 1

      The premium is not that small. The cheapest HGST is 21% more expensive than the cheapest Seagate I could find (4TB capacity)

    6. Re:HGST and Toshiba have been at the top for years by Anonymous Coward · · Score: 1

      Also, the deathstars were all the "IDE" consumer drives. The SCSI professional drives were quite good. It depended where the platters were sputtered and the electronics were made. I worked at the IBM facility in San Jose a few years before Hitachi bought the storage division...

    7. Re:HGST and Toshiba have been at the top for years by barc0001 · · Score: 1

      And how much is your peace of mind worth? Or your time spent cleaning up after that Seagate drive blows up on you?

      Good components cost more up front, but save you time and energy over their lifespan. I'm IT for a satellite office of a US multinational and unlike the parent company our workstations, dev servers and datacenter servers are all spec'd and scratchbuilt by me with decent components instead of being cheap and cheerful mass purchased crap from the 2nd lowest bidder. As a result my equipment failure rate is about 1/10th that of my US counterparts. It costs a little more up front, but the back end savings in time and productivity are definitely worth it.

    8. Re:HGST and Toshiba have been at the top for years by im_thatoneguy · · Score: 2

      Editor's note: In short: hard drives from HGST, a subsidiary of Western Digital, and Toshiba were far more reliable than those from Seagate

      I'm not sure that you can reach that conclusion from their data for the latest generation. The STD4000s were definitely hot garbage and the HGST 4TB were fantastic. But none of those drives are still on the market.

      The Seagate STD8000s they are running have a combined 1.2M drive days with 38 failiures. That's 1 average failure every 32k days. They also have around 1.75 failures per thousand.

      The WDC60 drives only have 40k days on them so they're still in the ballpark of similar performance. With 443 drives total again they are right around the 0/1 sample size. With such a small sample size the WDC is right on the edge of the margin of error on the sampling being larger than the failure rate of the Seagates.

      The new HGST 8TB generation only has 45 drives to sample from so absolutely nothing can be learned from it.

      I would say excluding the complete duds that were the ST4000s the Seagates are in their modern offerings equal to the WDCs.

      Honestly this last generation is looking so good for failure rates that I would just pick on capacity/price which Seagate is offering the best value for. Everything is a RAID these days and I don't have thousands of drives so my chances of replacing a drive is minimal and even then my only "risk" is the time/performance hit to rebuild the new drive.

    9. Re:HGST and Toshiba have been at the top for years by dgatwood · · Score: 1

      I'm not sure that you can reach that conclusion from their data for the latest generation. The STD4000s were definitely hot garbage and the HGST 4TB were fantastic. But none of those drives are still on the market.

      With an average of less than two months' use per drive, the current Seagate drives showed a 1.55% failure rate over three months, or a 6.2% expected annual failure rate, assuming all else is equal. With an average of five days' use per drive, the current HGST drives showed a 0% failure rate over three months, or a 0% expected annual failure rate. To be fair, the HGST cluster isn't nearly big enough and hasn't run nearly long enough for those numbers to be statistically signifiant yet. That said, in the times I've looked at their stats, no Seagate drive has ever done much better than about a 5% expected annual failure rate, and no HGST drive has ever exceeded one percent. So it isn't conclusive, but it is strongly suggestive.

      --

      Check out my sci-fi/humor trilogy at PatriotsBooks.

    10. Re:HGST and Toshiba have been at the top for years by Drakonblayde · · Score: 1

      They've said in the past that there are alot of factors that go into it. For example, the Seagates may fail more often, but they're also cheaper. And sometimes it comes down to simple availability. If they can't get the HGST drives in the capacity and quantity they want, well, they're a data storage company, it's not like they're *not* going to buy hard drives, even if they are more likely to fail.

    11. Re:HGST and Toshiba have been at the top for years by jedidiah · · Score: 1

      When I had to go into the hospital for an extended period I decided that the price advantage was no longer worthwhile. I was just replacing the buggers (Seagate) too often. I was no longer in the position to "baby" my arrays.

      So I dumped all my Seagate drives for WD.

      Those Amazon prices of yours really don't reflect a discount sufficient enough to justify the extra bother.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    12. Re:HGST and Toshiba have been at the top for years by guruevi · · Score: 1

      At these scales, a few percentage points of failure doesn't really matter if you save 25% cost. Even at relatively small densities of 100TB-1PB, 20% in disk cost savings is significant.

      Additionally, you should plan for failure anyway, disks fail, regardless of manufacturer. So buying a 'more reliable' drive is no guarantee and good backups are still a good idea.

      --
      Custom electronics and digital signage for your business: www.evcircuits.com
    13. Re:HGST and Toshiba have been at the top for years by tlhIngan · · Score: 1

      And how much is your peace of mind worth? Or your time spent cleaning up after that Seagate drive blows up on you?

      Good components cost more up front, but save you time and energy over their lifespan. I'm IT for a satellite office of a US multinational and unlike the parent company our workstations, dev servers and datacenter servers are all spec'd and scratchbuilt by me with decent components instead of being cheap and cheerful mass purchased crap from the 2nd lowest bidder. As a result my equipment failure rate is about 1/10th that of my US counterparts. It costs a little more up front, but the back end savings in time and productivity are definitely worth it.

      You know what I find to be the better predictor of drive longevity? The store I buy it from.

      I buy it online, and a lot of "computer" stores ship hard drives with little more than bubblewrap on it in the box. Or you buy a laptop hard drive, and it's shipped in a bubblewrap mailer. If you know how any shipping service works, you know that's not sufficient packaging to survive normal shipment.

      Even worse, buying multiple drives is often risky - some have the gall to bundle up the drives together as a massive block and then bubble wrap the entire thing. And it's no surprise a good chunk of them are *dead* when you get them. Some of the better shops cut up a multi-drive package and use the foam inserts, but then... that's it. It gets stuffed in the box with drives in the slots and left to fend for themselves. It's a surprise of all the times it's happened, one hasn't fallen out. (No, they're not taped down).

      Amazon I found had the best packaging - each drive was individually packaged in its own box with those plastic endcaps suspending the drive in the middle - what a drive manufacturer makes you use when you return a drive for warranty purposes. Even if you stuff multiple boxes in a bigger shipping box, the suspension keeps the drives from banging into each other. Basically, it's second to retail packaging (which is designed to really take a lot of abuse - not just in shipping, but when stocking and distribution at the store. It can fall from the shelf onto the floor while still protecting the drive inside).

      It doesn't matter if Seagates are more prone to failure if most of the damage to the drive happens in shipping from the store to you. You'd think people who work at companies that sell computer parts would be more careful, but no. Yes, they can take a lot of abuse, but really, they shouldn't have most of their life used up before the customer even gets their hands on it.

    14. Re:HGST and Toshiba have been at the top for years by barc0001 · · Score: 1

      I don't buy components for work from a mail order place, I have a hardware guy who runs his own company and he delivers after he picks it up from his reseller's warehouse in town - the same warehouses that supply all those mail order shops. When I get hard drives from him I am usually getting them in bulk (10+) and each are still in the styrofoam racking that he got from the warehouse when he picks them up.

      But for one offs or home use you are correct, better packaging means longer lived parts with moving components as a general rule.

    15. Re:HGST and Toshiba have been at the top for years by kinocho · · Score: 1

      Well, in my case it is because the first western digital I bought died on me three days after warranty expired. Admittedly it was just one disk, but I got so angry for that that I have never bought another disk from them. Until now I have been lucky. And as of late I have been purchasing 8tb disks from seagate.

  2. Comment removed by account_deleted · · Score: 1

    Comment removed based on user account deletion

  3. Re:hard drives from HGST ... far more reliable by HornWumpus · · Score: 1

    Tradition. This is the same company that shipped a warehouse full of known bad drives to market, as an accounting trick, back at the dawn of personal computers.

    The same name anyhow.

    --
    John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
  4. Re:hard drives from HGST ... far more reliable by whoever57 · · Score: 1

    There was a time when Seagate gave a much longer warranty than their competitors and the drives actually lasted longer.

    Sadly, those days are long gone.

    --
    The real "Libtards" are the Libertarians!
  5. Re:Bathtub model by known_coward_69 · · Score: 1

    Hitachi sells enterprise SAN's with their drives in them. Huge financial inducement to charge huge support fees and make your product better so as not to pay out.

  6. Re:Bathtub model by BenJeremy · · Score: 4, Informative

    I have a stack of dead 2TB+ Seagate drives sitting around here.

    Never again. They merged with Maxtor and the worst of both companies emerged out of the ooze.

  7. Seagate drives will have you singing... by Anonymous Coward · · Score: 5, Funny

    ...that old R.E.M. song "That's me on the hard drive, losing my partition"

    1. Re:Seagate drives will have you singing... by Anonymous Coward · · Score: 2, Funny

      I thought that I heard you crashing
      I thought that I heard a ding
      I think I thought I heard you dying

    2. Re:Seagate drives will have you singing... by arth1 · · Score: 2

      ...that old R.E.M. song "That's me on the hard drive, losing my partition"

      Back in the early 90s we joked "That's me on the Conner". I guess these days, few people know or remember Conner hard drives and their dubious reliability, so the update make sense! :)

    3. Re:Seagate drives will have you singing... by Agripa · · Score: 1

      And then later, Seagate doubled down and bought Maxtor.

  8. Re:hard drives from HGST ... far more reliable by technoid_ · · Score: 1

    Some of remember the old IBM Deathstar drives.

    https://en.wikipedia.org/wiki/...

    IBM sold their hard drive business and it has become HGST now. Weird how that is now reliable brand.

    --
    Two wrongs don't make a right, but 3 lefts do - Lew of GO magazine
  9. Re:hard drives from HGST ... far more reliable by Anonymous Coward · · Score: 1

    Perhaps you are thinking of Miniscribe, they shipped actual bricks in boxes prior to them going out of business.

    https://en.wikipedia.org/wiki/MiniScribe

  10. Re:hard drives from HGST ... far more reliable by HornWumpus · · Score: 4, Interesting

    Seagate. Back in 20 meg days. It was around the first time they went bankrupt.

    They got caught carrying a warehouse full of test failures as an asset. When caught by the auditors they doubled down and shipped them.

    --
    John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
  11. Re:hard drives from HGST ... far more reliable by __aaclcg7560 · · Score: 1

    I stopped buying Seagate drives years ago. They still suck?

    They did several years ago. Since then I replaced my Seagate drives with SSDs in the gaming PC and laptop, and Western Digital 1TB Red NAS drives in the file server. Although I did get a newer Seagate 3TB hard drive to serve as a backup drive for the file server. No problems with that drive yet.

  12. Re:Bathtub model by __aaclcg7560 · · Score: 1

    Please consider using SSD drives in your NAS. Results are optimal.

    As soon as a 1TB SSD becomes available for $50 each. I just replaced all my NAS hard drives last year. I'm hoping to replace those in five years with SSDs.

  13. Be sure, your RAID has a mixture by mi · · Score: 2

    For a single drive, go with the most reliable model. For a RAID, however, be sure to mix different manufacturers, models, and batches to avoid correlated failures...

    Because, if the failures are random, your mirror or even a large-count RAID5 will do fine for millennia, assuming you replace the failing ones in a reasonable time.

    But if the drives are all the same, they may all — after spending the same time in the same enclosure under the same load — fail for the same reason at the same time. Having hot-spares or multiple redundancy will not help you...

    --
    In Soviet Washington the swamp drains you.
    1. Re:Be sure, your RAID has a mixture by bobbied · · Score: 2

      Hot spare drives DO help, as long as you don't leave them spinning when they are in standby. It allows you to restore redundancy on a detectable failure and return your system to a "dual failure required for data loss" condition.

      Your point about not using drives from the same manufacturing run in a RAID array is somewhat valid, in that it can increase the possibility of having multiple drives fail at similar times, but if you *monitor* your drives, many failures are evident before they become catastrophic and you can usually be ready for recovery.

      However, the real issue with drive failures, in a RAID or alone, is having a backup strategy to recover from data loss events, which in my experience is *rarely* due to a drive failure, and usually due to user errors and environmental issues like power or cooling failures.

      My advice is to buy the best drives you can afford, back up offline and verify all your data regularly, closely monitor your hardware (power, cooling, drive errors) and take corrective action sooner rather than later. Don't run RAID for redundancy, only for performance if you can help it, and provision spare drives when using redundancy RAID arrays. Finally, BACKUP, BACKUP, BACKUP! ALWAYS BACKUP to offline storage...

      Spend your time wisely, backup often, don't argue about RAID, disk drive reliability and hot spares or not. Disk drives WILL fail, all of them, you WILL lose the data on them. Make sure you can deal with that problem first..

      --
      "File to fit, pound to insert, paint to match" - Aircraft Maintenance 101
    2. Re:Be sure, your RAID has a mixture by mi · · Score: 2

      Hot spare drives DO help, as long as you don't leave them spinning when they are in standby. It allows you to restore redundancy on a detectable failure and return your system to a "dual failure required for data loss" condition.

      Having a hot spare will cut your drive-replacement time by no more than a few hours. That's usually no more than a fraction of the actual rebuild time and, as my graph shows, that does not perceptibly affect the RAID's total MTTF. Moreover, if you monitor the drives — as you rightly suggest everybody should — then you can schedule the replacement of an about-to-fail one ahead of time, eliminating the difference between a hot and a cold spare entirely.

      You'd be better off using the slot occupied by the hot spare to increase the RAID's total capacity. Do have a cold spare — on the shelf — for the operator to slide in when needed, but a hot spare is simply not warranted.

      No disagreement on the importance of monitoring and backup-up...

      --
      In Soviet Washington the swamp drains you.
    3. Re:Be sure, your RAID has a mixture by bobbied · · Score: 1

      I think we are going to disagree about the hot standby, but in truth it boils down to what your response time to a drive failure is. I've had systems that I was responsible for that I couldn't lay hands on for at least 24 hours or more (They where in other countries). Granted it wasn't an ideal situation and we did have limited onsite support, I found that having a hot spare (as well as one or more on the shelf) to be useful to the MTBF of the system. My biggest problem was getting the replacement drive properly inserted into the correct slot, w/o causing a outage, not the hot standby recovery. Having the array already back into redundant condition before I tried to swap drives saved my bacon more than once. Pulling the wrong drive is way to easy. Of course we where shooting for Telco level reliability (99.999 or better) which basically means zero downtime, planned or not.

      If you are a 20 min drive away and have the spare sitting on the shelf, adding 20 min to the start of the recovery is not that much. Also, most of us really don't have fine nines as a real requirement, but something a bit less demanding. In my case, the uptime numbers where helped by the hot spare... Your mileage may vary.

      --
      "File to fit, pound to insert, paint to match" - Aircraft Maintenance 101
    4. Re:Be sure, your RAID has a mixture by Lady+Galadriel · · Score: 1

      Agree competely.

      Had a job where one time we installed a dozen or so low end RAID disk arrays. They had 8 slots but only 4 populated. So, RAID-5 was our only real choice, (given the application space requirement). The array vendor did supply a mixture of disk drive manufacturers, but it still did not help.

      We ended up with a lemon model, the IBM 3.5" 9GB SCSI. They started dying. Whence the root cause was determined, (bad batch of drives), we asked our array vendor to replace them all. Perhaps 40 more disks. They balked big time, so we threw our weight around, (the 800lbs gorilla gets action). Then we took a week to replace them, one per RAID-5 set at a time.

      To be fair, perhaps only 1/3 of the arrays had 2 of the faulty disks. None had 3 per array, (if I remember correctly). The rest got fixed with just one disk replacement.

      --
      Lady Galadriel
    5. Re:Be sure, your RAID has a mixture by mi · · Score: 1
      If:
      • I couldn't lay hands on for at least 24 hours or more and
      • you have a spare slot in your RAID-enclosure anyway

      then yeah, you may as well put the spare drive into the otherwise empty slot instead of on the shelf.

      --
      In Soviet Washington the swamp drains you.
    6. Re:Be sure, your RAID has a mixture by guruevi · · Score: 1

      Large count RAID5 WILL NOT DO FINE (trust me). Your RAID5 takes exponentially more time for each drive you add and during that time your data is in a RAID0 situation. RAID with at least 2 parity drives is the minimum requirement, regular mirrors if you have failover systems or triple mirrors if you're in a SAN-situation.

      --
      Custom electronics and digital signage for your business: www.evcircuits.com
    7. Re:Be sure, your RAID has a mixture by mi · · Score: 1

      Large count RAID5 WILL NOT DO FINE (trust me).

      The larger the count, the bigger the risk, yes. And yet, if the drives are all different, it will still do fine for thousands of years. No, I will not just "trust you". You may have some personal anecdote to "prove" it, but the math I referred to speaks for itself.

      RAID with at least 2 parity drives is the minimum requirement

      Wasteful bullshit. All too common among Infrastructure people (who never even studied Statistics, much less got a decent grade), but still bullshit.

      In all likelihood, you are populating your SAN-enclosures with identical disks taken from the same box. That is what you ought to stop or no amount of mirroring will help.

      --
      In Soviet Washington the swamp drains you.
    8. Re:Be sure, your RAID has a mixture by guruevi · · Score: 1

      Even a single anecdote would disprove your theory of 'thousands of years'. There is no such thing as 'thousands of years' of runtime on a drive, you're talking MEAN time BETWEEN failures (or MTTDL, mean time to data loss) and even then you have to account for all the drive configurations in existence, in an ideal world.

      You can do the calculations, go ahead, there are calculators on the Internet for you. There used to be an Excel spreadsheet from a Sun engineer a long time ago, but

      I'm sure you won't understand the content of this article, but for reference to other people: http://queue.acm.org/detail.cf...

      --
      Custom electronics and digital signage for your business: www.evcircuits.com
    9. Re:Be sure, your RAID has a mixture by mi · · Score: 1

      Even a single anecdote would disprove your theory of 'thousands of years'.

      It is not a "theory", I offer a mathematical proof. You, on the other hand, would not offer even an anecdote.

      There is no such thing as 'thousands of years' of runtime on a drive,

      Not on a drive, but for an array — a RAID5 with disks failing randomly will survive for millennia before two unrelated failures happen within the period of replace/rebuild-time of each other.

      I'm sure you won't understand the content of this article

      Ah, how insulting... I don't think, you understood it :) They take their fancy charts from a NetApp's "study":

      The NetApp comparison is not specific about the bit error rates of the devices tested, the reliability of the drives themselves, or the length of the period over which the probability of data loss is calculated; therefore, we did not attempt to reproduce these specific results. The important point to observe in figure 1 is the stark measured difference in the probability of data loss between RAID-5 and RAID-6.

      That's a rather uncritical acceptance of data supplied by the vendor of storage solutions obviously interested in selling bigger boxes and more of them.

      And, no doubt, they too populated their boxes with identical drives leading to correlated failures. Do not do that.

      --
      In Soviet Washington the swamp drains you.
    10. Re:Be sure, your RAID has a mixture by guruevi · · Score: 1

      Again, you don't understand the mathematics involved, linking to your own (bad) calculation is not proof. This calculator explains it much better than I can: http://wintelguy.com/raidmttdl...

      RAID5: 10 drive groups of 8 drives, 6TB drives:

      MTTDL: 1.2years
      Probability of data loss is more than 50% in a year
      MTTDL due to multiple disk failures: 286 years
      Do you understand how incredibly low a MTTDL of 286 years is?

      Just swapping it over to RAID6 with the same settings and the probability of data loss is 1% over 10 years (which imho is getting close to being unacceptable)

      There are a number of other calculators and some really good explanations on how to make them, there is no situation where RAID5 does any better using real world numbers than RAID6 or RAID10.

      Correlated failure is a figment of your imagination. I'm running several arrays of several 100s of TB, close to 500 drives in total, I don't think I ever bought more than 24 from the same type or vendor, yet, I've seen failures of a Seagate and a Hitachi in the same array in the same night, I don't see correlated failures (where multiple disks of the same model and same age will fail at the same time) once the disks have passed my QA. I see trends where groups of drives will fail around the 1y, 3y and 5y mark but that's a correlation of age, not by manufacturer.

      Google has some interesting statistics as well, you'll see the same cluster around age of the drive being much more an indicator for failure than temperature swings and whatever else Google collects data on.

      --
      Custom electronics and digital signage for your business: www.evcircuits.com
    11. Re:Be sure, your RAID has a mixture by mi · · Score: 1

      linking to your own (bad) calculation is not proof

      It is proof, unless an error is identified. Math has this nice property about it, that it is not subject to opinion. What is bad about them?

      This calculator explains it much better than I can

      Because you do not understand the Math...

      there is no situation where RAID5 does any better using real world numbers than RAID6 or RAID10.

      I never claimed, that RAID5 does better. My claim is, it is perfectly sufficient and that the higher redundancy does not improve reliability by high enough margin to warrant sacrificing storage capacity. RAID5 guards against random failures. I argue, that trying to guard against correlated ones is futile and your own anecdotes confirm that.

      I've seen failures of a Seagate and a Hitachi in the same array in the same night

      Could happen too — for example, when ventilation fails, etc.

      Do note, that the multiple levels of redundancy did not save you. I'm sure, you dealt with the multiple simultaneous failures somehow and point out, those fall-back mechanisms are what saved you, not the extra redundancy. It buys you vanishingly little...

      I see trends where groups of drives will fail around the 1y, 3y and 5y mark but that's a correlation of age, not by manufacturer

      Whatever the source of correlation, having additional redundancy is not helpful — when drives fail together, it is due to external factor(s) and you still need other mechanisms in place to deal with that.

      Correlated failure is a figment of your imagination.

      Dude, you've enumerated several anecdotes confirming correlated failures being real.

      you'll see the same cluster around age of the drive being much more an indicator for failure

      Wish you'd included the link... But, hey, that may be because makers source components (like motors) from common suppliers...

      --
      In Soviet Washington the swamp drains you.
    12. Re:Be sure, your RAID has a mixture by guruevi · · Score: 1

      Okay, quick scanning of your site:

      The entire section of Poisson, having more drives does not make it less likely that one will fail.
      "Failures occur every mttf hours" - that's not what MTTF means. MTTF simply means that if you have a collection of "n" drives, that you can statistically expect 1 failure every (MTTF / n). So if you have a manufacturer says, your MTTF is 100 years (which is about what they promise), then you will have a failure amongst a set of ~1M drives every 1 hour.

      Poisson's distribution simply calculates the chance at multiple failures which is easy to calculate:
      If you only had 1 drive: 0 or 1 drive failure per 100 years of runtime: 36%; 2 failures: 18% etc.
      If you have 2 drives in a set, all MTTF implies is that the chances remain the same but are now per 50 years of runtime.
      If you have 20 drives in a set, the same per 5 years of runtime

      Even IF your calculations are correct, you're getting a MTTF of the array in the 100-200 year range, that implies that if, in the world there are 100-200 arrays set up like that, 1 will fail every year. Even a small datacenter has the capacity to hold that, so you find it acceptable that an entire array loses all it's data once per year on average? And you haven't even accounted for the rate of unrecoverable errors which is ~1 every 10-12TB. Good luck with that data though.

      --
      Custom electronics and digital signage for your business: www.evcircuits.com
  14. Re:hard drives from HGST ... far more reliable by Luthair · · Score: 1

    Because they don't. Backblaze takes consumer drives, plonks them in an environment they weren't designed for (servers) and runs them 24/7. Hardly surprising drives have an abnormal failure rate.

  15. Re:hard drives from HGST ... far more reliable by fnj · · Score: 2

    Corporate apologist shill alert. No one falls for that that old lie any more. The difference between "consumer" and "enterprise" drives is a different label and a huge price gouge. Nothing more. Actually 24x7 operation is a lot easier on them than constant cycling on/off.

  16. Re:hard drives from HGST ... far more reliable by alvinrod · · Score: 1

    Yeah, I remember when Seagate had the most reliable drives and I used them for many years until their quality went way down. I'm not buying spinning disks any longer, though the last ones I bought were Samsung manufactured (funnily enough they sold their hard drive business to Seagate several years ago) and some of those are still seeing regular use in some older boxes or for local backup. By the time they fail I'll probably be able to replace them with an SSD that's just as large, but costs less than what I paid for those spinning drives when I bought them new.

  17. Re:hard drives from HGST ... far more reliable by Archangel+Michael · · Score: 2

    BackBlaze already did this in an episode. They have proven that there is no statistical difference between Consumer or Enterprise grade. In fact, IIRC, the Enterprise Drives were slightly worse (but within statistical variance).

    The reason why Enterprise are made, is for PHBs who see "Enterprise" and think ... "BETTER!!!!"

    The sad thing is, that kind of crap marketing actually works.

    --
    Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
  18. Summary of the summary by DontBeAMoran · · Score: 1

    Seagate is not reliable.

    --
    #DeleteFacebook
  19. Re:hard drives from HGST ... far more reliable by bobbied · · Score: 2

    Tradition. This is the same company that shipped a warehouse full of known bad drives to market, as an accounting trick, back at the dawn of personal computers.

    Wow, I worked for a company that boxed and "shipped" a pile of equipment to a host of resellers who hadn't ordered anything. The shipping terms where FOB-Origin so the resellers where responsible for picking up the equipment, which of course they didn't, because they didn't order it. Then, when the resellers got their invoices and complained that they didn't order all this stuff it got "returned" for a refund. Of course the "sell" date was in one quarter and the "returns" happened in the following one. It was just a way to realize a pile of sales and meet the quarter's numbers. The FTC got wind of this and the CFO got spanked and the company got fined, but it took a few years.

    I didn't think it was a common practice....

    --
    "File to fit, pound to insert, paint to match" - Aircraft Maintenance 101
  20. Apple for the win by DontBeAMoran · · Score: 4, Funny

    It's sad that Windows and Linux users have to go to such troubles.

    Me? I only buy Macs. Because Apple takes the 1% of the best drives made in each manufacturing lot and puts them in their Macs. That's why Macs are so expensive.

    I mean, this has to be the reason, right? Surely they're not just buying the same parts as Dell and others and just selling overpriced computers and pocketing the profits.

    --
    #DeleteFacebook
    1. Re:Apple for the win by jedidiah · · Score: 1

      > It's sad that Windows and Linux users have to go to such troubles.

      Why do you have this deranged idea that PC users are forced to buy any particular brand of hard drive? We can buy any brand we like.

      --
      A Pirate and a Puritan look the same on a balance sheet.
    2. Re:Apple for the win by DontBeAMoran · · Score: 1

      My joke was that you had to do any research at all and that Apple, being costly, had the best 1% of hardware and PC peasants had to buy the 99% rejects.

      --
      #DeleteFacebook
  21. Re:Bathtub model by DidgetMaster · · Score: 1

    Good luck with that. The decline in SSD prices seems to have slowed to a crawl. I bought a 250GB SSD for about $75 a few years ago and they seem to be about the same price now. I now wonder if we will ever see TB SSDs for $50.

  22. Re:hard drives from HGST ... far more reliable by TheRaven64 · · Score: 1

    I'm not sure how they calculated the failure rates. For one model, they had 5/400 fail, for another 13/157. These are both pretty high (though with such small number, could be statistical flukes), but neither is 30%.

    --
    I am TheRaven on Soylent News
  23. Re:Bathtub model by TheRaven64 · · Score: 2

    You and I have very different definitions of optimal. 2TB SSDs cost a little under double what 10TB hard disks cost. For a consumer NAS, the performance difference vs having a reasonable amount of RAM / L2ARC is negligible over the network (the disks are not the bottleneck), but the cost for the same capacity is at least 10 times (probably more, as you need a lot more SSDs and so need more space, more controllers, and so on).

    --
    I am TheRaven on Soylent News
  24. Re:hard drives from HGST ... far more reliable by __aaclcg7560 · · Score: 1

    Maybe creimer has a hard drive in his head too?

    As a rule of thumb, I don't bother to memorize stuff. Specialized knowledge should always be documented. If not documented, I write the documentation and stick it into a knowledge base.

  25. Hard disks ? by krouic · · Score: 1

    Hard disks are so old fashioned, why don't they use the Cloud ?

  26. Re:Bathtub model by rajafarian · · Score: 1

    There was a time when I would have considered a Seagate drive but when they bought Maxtor, I swore never again. I don't think I would buy one even for someone else's computer.

  27. Re:Bathtub model by jedidiah · · Score: 1

    A NAS isn't nearly fast enough for this to even matter. The N part of the whole thing will be your big bottleneck. Even the random read/write access pattern where SSD shines will be trashed by the fact that you're connecting across the network.

    A single piece of spinning rust can saturate gigabit on it's own.

    --
    A Pirate and a Puritan look the same on a balance sheet.
  28. Re: HGST and Toshiba have been at the top for year by Brockmire · · Score: 1

    Buy on sales. Bought several from newegg.ca when they go on for less than an Ironwolf.

  29. Re: Three kinds of shit. by Brockmire · · Score: 1

    Three? I've seen a poster detailing at least 10 types of shit.

  30. Buy HGST right away by Applehu+Akbar · · Score: 1

    BackBlaze is about to put in a behemothic order as it gets ready to take on all the CrashPlan customers.

  31. Re:hard drives from HGST ... far more reliable by HornWumpus · · Score: 1

    'Stuffing the channel' is common enough to have a cute name. In very limited cases, it's even legal. But consult a shyster.

    --
    John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
  32. WTF? by Artem+S.+Tashkinov · · Score: 1

    Seagate stats don't make any sense: ST4000DM001 - 400 - 5 - makes it 1.25% failure rate - I see 30.43% in the table. Likewise with ST4000DX000.

    Could anyone explain how the f they calculated Seagate data?

    1. Re:WTF? by mattventura · · Score: 1

      They extrapolate annual failure rates. You divide the number of failures by the total amount of time the drive has been in service (sum of all drives - i.e. 100 drives for 50 days = 5000 drive-days), then multiply by 1 year. In this case, those 400 drives have been in service for a total of 6k days (average of 15 days/drive). If 1.25% of drives crap out in 15 days, you can probably see how that becomes a 30% annual failure rate. Granted, there's some bathtub curve going on here, so that particular statistic isn't very reliable - it may as well be interpreted as "this drive has a 1.25% chance of being DOA"

  33. Re: HGST and Toshiba have been at the top for year by scdeimos · · Score: 1

    No legit vendor would do this because passing off referb parts as new in most western countries is completely illegal.

    Only if they're sold as "new." It's generally legal to resell them so long as you indicate that they're refurbished products.

  34. Re:hard drives from HGST ... far more reliable by mattventura · · Score: 1

    It makes sense. They extrapolate an annual failure rate, since failure rates over time need some kind of time period to be useful. Using your example, if 5 drives fail over 5998 total days of drive time, that means any given drive has a 30% chance to fail in a given one-year period. As an example, if there were 50 drives in that sample, that would mean they ran those 50 drives for 120 days, and in that time 10% of them died, which is terrible reliability.

  35. Re:hard drives from HGST ... far more reliable by Woeful+Countenance · · Score: 1

    They give an example involving 15 drives whereby two failures in 100 days gives a daily failure rate of 2%. (Two percent of what?)

    Good question. If two failures in 15 drives in 100 days = 2%, what would be the failure rate if two drives out of, say, 150 drives failed in 100 days? What if three drives failed in 101 days?