Follow Slashdot stories on Twitter

 


Forgot your password?
Close
typodupeerror
Hardware Hacking

Journal tomhudson's Journal: Seagate ate *another* RAID 1 33

Journal by tomhudson

Yes, 2 more Seagate drives, 2 more failures. So now I have 4 bad Seagate drives.

I figured that, since Seagate was going to be half-way decent (after a bit of haggling) over the two new drives that failed, I'd use the long weekend to try out openSUSE 10.3 - so I bought 2 more Seagate drives, exact same make and model, from a different supplier.

An hour ago, one of the drives started making funny noises ... you know the type of noise that a drive makes when it keeps seeking over and over ... just like the first set.

Seagate 320 gig hard drives - ST320620A - avoid them like the plague!

Its pretty bad when I have to depend on the out-of-warranty +10,000 hours operating time Western Digital 250 gig that's been running in that same box for the last 3 years, because 4 brand new Seagate drives are just so much crap!

Well, look on the bright side ... at least I know now that it wasn't an isolated incident. Better to find out now than in 6 months, when they would have been loaded with data.

Still, what a way to waste 2 days of work on a long weekend :-(

This discussion has been archived. No new comments can be posted.

Seagate ate *another* RAID 1 More

Seagate ate *another* RAID 1

Comments Filter:
  • 1. Discover entire bogus/broke-down run/series of Seagate drives

    2. Originate/Instigate class-action lawsuit for lost time and/or data.

    3. PROFIT!!!!

    (4. Remember your friends)

    I'd say something about a beowulf cluster-f of these things, but we've all seen trash heaps before...

    • What are the odds that 4 out of 4 drives would fail, from two different retailers? The only thing they have in common is the make and model. Its certainly not my computer, because the Western Digital is still chugging along, with over 10,000 hours on it.

      At this point, I just want my money back ... I definitely *don't* want any more Seagate drives.

      • What are the odds that 4 out of 4 drives would fail, from two different retailers? The only thing they have in common is the make and model. One reason why I advise my students (apart from the 'raid is not backup' and 'raid 0 is not raid') to REALLY consider purchasing several disks of the same batch. Seagate is a fabulous brand, but a faulty batch is a faulty batch (if I'm not mistaken there also was a bit of an issue with the 250 gigs in the last g4 imacs).

        But the waste of time really is a shame. (having

        • I *tried* to purchase different batches, but from the model numbers, it looks like they were made on the same day. If it were a car, I'd say it was made on a Monday ...

          I could have bought a cheap new tower for what I've spend on drives, or for $50 more, a new laptop.

          I brought back the second set of drives to the other retailer Friday - now I remember why I haven't dealt with them in 5 years. Their only answer was to run the extended SMART diagnostics (and come back in 4 hours). They're a Windws-only sh

          • Good grief... well, you tried hard enough to get it to work! I would've given up earlier probably, and contacted Seagate from the start. A lot of retailers aren't worth spending time on, and start looking glazed when you mention 'linux', 'apple' or even 'partition'.
  • Wow. I had already decided not to buy from Seagate, as I mentioned in replies to previous (Tom's) journal entries. Now I'm really sure.

    On a slightly more than tangentially related note, I've got a Samsung 120 GB SATA drive that has been having major problems. I bought it in early September, and since then, I've been having all kinds of problems with files getting corrupted and having to run chkdsk, including a total disaster that forced me to reinstall Windoze from scratch. I have discovered that m

    • If you can boot off a linux cd-rom,

      1. choose "rescue"
      2. log in as root at the prompt (no password needed)
      3. smartctl /dev/hda -a (or for the latest releases, /dev/sda -a).

      You'll see stuff like the number of errors, etc.

      Google's studies show that the thing to look for is the raw read error rate and the number of "misses". All the drives that failed have raw read error rates in the hundreds of millions (one has over a billion), which is bad for drives that have, in 2 cases, 1 day of use.

      Noise (rattling, hi

    • Re: (Score:1)

      by trum4n ( 982031 )
      i have 7 maxtors in this box. the oldest is my windoze drive, 5 years, 20gb. the newest is a pair of 250 SATA's in RAID 0. This box flies, and is perfectly stable. The rule is, put a drive you dont are about as C: with windoze on it. If it fails, who cares, its just windose, u have to install it every 3 months anyway. All teh programs and media go on good drives. The biggest reason i do this is because 99% of viruses cant disk span. They are trapped on the windose drive, protecting your media.

      • All teh programs and media go on good drives. The biggest reason i do this is because 99% of viruses cant disk span. They are trapped on the windose drive, protecting your media.

        There's something I'm not understanding here. What good does it do to install the programs on other drives? Once you lose the Windoze drive, you lose the registry, and so even though the executables and support files are all on the other drives, you still have to reinstall everything, or Windoze won't recognize those programs. A

        • Re: (Score:1)

          by trum4n ( 982031 )
          "And even though a registry backup might partially solve that problem" That's the answer, and also, the more important things to me is the support files. Like game saves and documents. I have all the software on CDs, i don't care about the programs.
  • Sounds like something's fishy with that particular model. I have a Seagate 500GB SATA/300 (7200 RPM, 16MB cache) that I bought at the start of summer, zero problems, model ST3500641AS-RK. Says "Product of Thailand".

    I remember a few years ago losing all three new 60GB IBM DeskStars within 40 days of purchase. They were popularly known as IBM DeathStars, and word was all over the web that the DeathStars made in Hungary were crap (over 60% failure rates!) Perhaps yours were made in the "wrong" factory, o

    • I should have mentioned they were all "Made In China", like pretty much everything else nowadays. the only reason I got the 320s instead of the 500s was because they were out of 500s. Just my (bad) luck.
      • SEAGATE:
        I've still got two 160's, one in China, one in Singapore... both working ADMIRABLY, in fact they often outperform my Maxtor/WD's in speed both on writes and 'moves/copies'...
        They're WAY out of warranty...

        MAXTOR:
        I have on the other hand, had 3 Maxtors die on me, (got refurbished warranty replacements that are working fine... strange that.)

        WESTERN DIGITAL:
        I've yet to have problems with WD's,

        IBM:
        I've had two IBM Deskstar GXP65's and one GXP75 die on me within months of receiving them.
        Speaking of IBM dr

        • I had an old 80 meg Western Digital that, after years and years of running, I stopped using the machine it was in (286/20). Just for the heck of it,I booted it up a few years ago before junking it, and it still ran fine.

          I had an old Seagate (340 meg) in a 386 that never caused any problems.

          Maxtor, on the other hand, sometimes is DOA. The place I used to work at, it was always the Maxtors that I would RMA.

          Western Digital - my home box - I've had one 160 gig fail (its not "really" dead, just started to

          • Yeah, I've been using server cases or at least very large cases with lots of big (slower) fans lately... they're cooler, they last much longer than smaller fans (which start to rattle after awhile).

            As for hard drives. I had a big maxtor die on me, it was my backup drive in college. I must've lost a good 100 gigs of media and at least another 70 gigs of random miscellany, some comp sci work, some god only knows what it may have been...

            We call that lesson two things... "expensive" and "learned" :)

            Started ba

            • I've got most everything backed up. The most important stuff was several years worth of email, which I tarballed and stuck on a 2 gig usb key.

              funny how, in the great scheme of things, the one thing I would absolutely not want to lose is my email. Code I can always replicate (and probably improve at the same time), but email ...

  • Sounds like you've definitely had a horror run with Seagate drives! I'd definitely be looking for another manufacturer, that's for sure.

    OTOH I've had a pretty good run with them. I have 6 320GB Seagates, all about 2 years old and not one problem with them. I also have 8 750GB Seagates and a few smaller Seagate disks.

    It could be that my disks have been sourced from different factories than yours.

    • From the other comments posted, it may be because they're now making them in China instead of Thailand, Malaysia, etc. Not one of the 4 has lasted more than a day without starting to give problems ... that's seriously screwed up.

      Even though they wre bought at 2 different stores (and not the same company), the serial numbers tell a different story:

      1. 6QF14MNF
      2. 6Q1F4ZCY
      3. 6Q1F47FY
      4. 6Q1F58MN
        1. Assuming they use the 34-base system (0123456789 ABCDEFGH JKLMN PQRSTUVWXYZ), there are only 58MN - 47FY between them.

          f

      • From the other comments posted, it may be because they're now making them in China instead of Thailand, Malaysia, etc. Not one of the 4 has lasted more than a day without starting to give problems ... that's seriously screwed up.

        A year or two ago, I had a serious run of bad Maxtor drives - 'bad' as in 'entire room full of machines losing HDDs in a few months'. (To add insult to injury, their 'warranty' system entailed international mail, at our expense; as it happens for these drives, that cost only a tin

      • Re: (Score:1)

        by arb ( 452787 )
        (Hmmm... I'm sure I posted a reply to this, but Slashdot seems to have eaten it...)

        The last digit in the serial numbers could be a check-digit, meaning the disks are even closer together. Even though you purchased them from different retailers, the two retailers most likely get them from the same distributor. There is a good chance my disks come from a different batch/factory.

        When I get a chance to open up my machines I will try to find out where my disks were made - that shouldn't be too far off, I have to

        • I hadn't thought of the check digit. If the last one is a check digit, they're only 4,887 units apart.

          These were made in China. As another poster pointed out, its probably Maxtors' old factory. I knew that Seagate had bought Maxtor, but I didn't think they were so stupid as to try to get away with rebadging Maxtor drives. None of the Maxtor drives over 40 gig have ever lasted to the end of their 3-year warranty period. To expect these to go 5 years is unrealistic.

  • I wonder if it's just the 320s. In the past year, I've had only two Seagates (one 750GB, one 250GB) fail out of the box, out of a total purchase of something on the order of 150 mixed between 250GB, 500GB, 750GB, and 1TB drives.

    At least their warranty repair service is relatively hassle-free.

    • It may be just that particular model, since some of the other drive sizes may have been sourced from a different factory. Its just my bad luck. I had wanted to pick up a couple of 500 gig drives, but the first retailer didn't have any more in stock (they had sold the last two that morning), so I ended up taking the smaller drives.

      My original plan had been to buy two 500 gigs, install them as a raid1, and after a month, buy 2 more 500 gig drives as a second raid1. This way, I would have /, /usr, /var, /et

      • Re: (Score:2)

        by Reziac ( 43301 ) *
        When you have a rapid series of HD failures like that, look *first* at the RAID controller, not the HDs. The HDs may well be bad, but I've heard of too many cases of RAID going berserk. And considering the issues between large HDs and SATA controllers, there's some question in my mind as to whether RAID controllers universally support the HDs you'd expect them to.

        And I've seen the results of one case firsthand that was blamed as a HD failure, but since the data was randomly striped (many files had parts of

        • I wouldn't use a hardware raid controller - they're too limiting in what you can do, esp. in terms of recovery when you're running raid5 and the controller/motherboard dies, and the manufacturer is out of business (saw a scenario like that a month ago). Also, these are pata, not sata. The WD on the same controller as one drive is still working, and so is the dvd on the other controller. Its definitely the drives. When they make funny noises after less than a days' operation, and smartclt shows an ever-incr

          • Re: (Score:1)

            by Gr8Apes ( 679165 )
            That was definitely a batch of bad seagates. If China keeps putting out sub-par products like they have lately, I'll make sure that whatever I buy is NOT made in China, at least until they get back on track.

            I personally like seagates, and have had very little trouble with them over the years. Fujitsu was another that had near zero problems, although that was a much smaller data set. Maxtors do tend to fail, and are the drive with the largest number of failures. WD is too noisy, and generally too hot, althou
  • This does *not* encourage me about the replacement drive I'm being mailed from Seagate right now...
    • Me, either. And here I was getting ready to build a nice big RAID array with those exact drives. I was looking at the 500GBers as alternatives (for a somewhat larger array). I think I might go that route.
  • I was never interested in the 320GB models, something about it not being a multiple of 100 was unnerving to me. I have 4 Seagate 400GB drives, 1 is IDE and 3 years old now, the others are SATA and about a year old. They run great. However, I did notice that on all 3 Seagate 160GB drives I tried that they would be wisper quiet at first and after 48 hours of run-time they would then begin making audible clicking noises out of nowhere.
    I also have 2 Maxtor 40GB drives that are 5 years old now and despite be

    • I would settle for audible clicking noises at this point :-)

      I thought the 320 was a weird number also. If the first retailer had had 500s in stock, I'd have bought them. Heck, for what I've spent for 4 defective drives, I could have bought a couple of 750 gig drives, or thrown in another $50 and bought a new cheap laptop or desktop.

      To say I'm disappointed is an understatement.

Slashdot Top Deals

Swap read error. You lose your mind.

Close