08-23-2011 12:16 PM
I bought 5 barracuda XT 3tb drives a couple of months ago to use on my storage server. It uses an Adaptec 3805 sata/sas host controller card, which is pretty good.
The array comes in at 11tb and worked fine for the first few days. It is blisteringly quick and I was all smiles... then the array degraded. 1 drive failed, and though the controller always rebuilt it successfully it would fail again after a day or so. It seemed some type of intermittent problems as after the rebuild the drive would happily function as though nothing had occured. The same drive failed about 8 times in just under 2 weeks though I ran Seatools on all the drives, they all passed all the tests. They all have the lates firmware CC44.
1 actually did have another drive fail twice, and another fail one once. Just to be sure it wasn't my controller card I swapped it out with another one I have an adaptec 51245. The same drive failed after a day, so I figured time to go back. .
Anyway, I sent the drive back. Seagate replaced it no problem and the new drive has been working fine since. However, after two weeks a second drive (the one that originally failed twice) has now failed 5 times with each failure period happening more quickly. I haven't run seatools on it but I expect it will be reported ok just as all the drives were initially, even when they were clearly failing. I also expect the same thing will happen sooner rather than later with the drive that has only failed once.
At any rate, I will run seatools then I'm sending back the second drive tomorrow. Essentially I now have 3 disks that have worked without failing only 2 of which came from the original batch (one was a replaced by seagate) but that means of my batch of five, three of them are faulty. This cannot be good.
I am scared to even turn the server on now as another happens to fail before the replacement i'm stuffed. It took me ages to back up the stuff to tape and some of the new content hasn't even been backed up at all. I'm scared to even attempt a backup run in the degraded state just incase another drive fails before I get the array rebuilt. In the meantime I have ordered a Hitachi desktar7k3000 3tb as a spare. I dare not order another Barracuda XT.
I am wondering if anyone else has had problems with drive failure rates on this model.
Solved! Go to Solution.
08-23-2011 02:38 PM
Not ure what to say tbh, I have one but it's not a raid set up I use it in, it's my boot drive and it's still 100% 3-4 mths on, I have found an article that says raid can kill drives, but I can't find it because it was buried in a thread I was reading elsewhere, any chance you can get some temperature readings from them, the knows suggest a drive works best hot, I don't agree myself, I think they are like cpu's and need to be kept cool.
08-23-2011 08:18 PM
08-24-2011 02:21 AM
Yes drives generally need to be kept cool, and this particular host has a all aluminium heat sunk 8 drive caddy (the five 3tb drives in it 3 spare slots). I has 3x 50cm fans on top and an additional 120 internal fan that feeds air through the caddy. The drives generally run hot in 'stand alone mode' but this caddy keeps them considerably cooler. They are fairly warm to the touch mounted as opposed to particularly hot which is how they run outside the caddy.
And yes quite right, I disconnect them from the raid controller to run the seatools software.
I'm just surprised that these drives seem so failure bound. As a rule I only buy Seagate drives (except when I can't source something from my suppliers) and I have some 45 Seagate drives in use at home at the moment, and a box full of others that have been replaced as capacities have grown. In ten some years I've only sent back 2 drives. I couldn't even hazard a guess at how many have passed through my hands.
I do think your quite right about raid arrays being more demanding as the two drives that failed were barracuda 250gb and 500gb respecitively that ran off the adaptec 51245 on my main pc. However this failures happend 3 and 2 years ago and the replaced drives have never blinked since. Also they failed after some time of being in use (at least a year) not like these fresh of the production line with latest firmware models. It currenly runs 12 x 500 7200.12 barracuda in raid five and it does this all day long without so much as a hiccup.
I will try the tools you have suggested fakbar, and also see how I get along withthe Hitachi which should be delivered today.
08-24-2011 07:28 AM
try the more expensive constellation drives and see how much better they work
08-24-2011 10:27 AM
Hmmm, that isn't really the issue though is it. The drives perfrom fantastically for my purposes when they work, the array is very quick indeed. The trouble is not one of performance but reliablity.
08-31-2011 05:49 AM
Ok sent back the second XT drive that was giving me problems. Got my brand new Hitachi drive, and the Array controller went about it's business integrating the drive into the array which it did quite happily. That was about 5 days ago and so far everything has been ticking over nicely. So! it appears my second XT drive was definately faulty (yes I just was never 100% sure as I've never had 2 drives in a batch of only five fail).
Now my array consists of 4 3tb XT's and one Hitachi, but it doesn't seem to mind. Waiting for Seagate to send me the replacement XT. Taking longer than the first time, by this time with the first drive I had already had notification that the drive had been dispatched. This time I've had only notice its been recieved, that was some 4 working days ago.
10-19-2011 10:43 AM
Just received my 4th drive from Seagate today. Oh I looking at my previous post I forgot to mention, I did receive the third replacement from seagate and guess what! The drive didn't even make it into the array before it failed. In fact the 3rd replacment arrived already faulty. I have one of these sata docking stations that you can plug the drive directly into - which I did as soon as I unpacked it. I wanted to check the firmware and run seatools on it etc. As soon as I docked it I knew something was up because my whole desk started vibrating like a belt sander. I Ran Seatools anyway but the drive failed in just a couple of seconds with a 6E616D8 error code and a bunch of 'Bad LBA' errors I thought would scroll into next week.
Anyhoo, this time phoned Seagate as I was a bit miffed. Spoke to a guy from tech-sup in the US (don't ask me why they patched me through there), He was very sympathentic and handled the the RMA for me. DHL came a picked up the drive next day. The return label I printed was for the Netherlands, (go figure) as I normally send them locally (UK), but at least this time I didn't have to pay £5 for the return cost.
That was on teh 29/9/11 low and behold the replacement drive arrived today some 20 days later. Thankfully the array was fine had been working without issue, then, believe it or not no more than an hour after the postman's delivering the package, the array failed! Yes you guessed it, Seagate drive number 4, failed.
The array is rebuilding at the moment so I won't be able to do anything until it's finished. The plan was to replace the Hitachi and just have a homology of seagates but that may not be such a good idea as statistically the Hitachi is the most reliable device in the array.
This is the second time this drive (number 4) has failed, and though it is possible it's just a bad sector or something,these drives seem to have increasing failure rates once initial problems occur.
I wonder if I can get a refund on these drives and just replace the whole batch with Hitachi's. I emailed Seagate registering my dissatisfaction about a month ago but I've not had a response.... woe is me.
10-19-2011 04:09 PM - edited 10-19-2011 04:09 PM
I still think the array is breaking them Jim, I've no idea how one gets to sound like a belt sander tbh, it looks like you are chasing a refund right now, not sure how you will go with that myself, but if it's what you want good luck with it, I have to admit at nearly a grand's worth of drives I'd be feeling that way also.
10-20-2011 09:45 PM
Nope, definately not the array. It's the drives that are simply failing. Like I said, the 'belt sander' replacement drive came from seagate and didnt' even make it to the array. It went straight from the packaging to the USB disk caddy where it failed immediately. Ah well such is life, I'll just avoid these drives like the plague in the future.
10-21-2011 04:05 AM
The dock might not be compatible for a drive this size, and make sure you aren't daisy chaining lots of PC stuff off one wall socket, this will cause issues if the socket can't provide steady flow due to excessive demand.
10-21-2011 09:04 AM
you mean the Docker caused the drive to fail? Interesting hypothesis. Would it also cause the drive to spin in a off center/unblalanced manner to causes excessive vibrations?
Daisy chaining wall sockets, another interesting hypothesis. You mean such that the current demands of the appliances exceed the wall socket current rating, wouldn't that cause the fuse/breaker to trip?
10-21-2011 02:06 PM
10-21-2011 04:16 PM
Not exactly what I mean't Jim, the dock may not support a 3TB drive fully, most that are over 2 years old will only be certified up to 2TB unless stated on the box, also if the dock wasn't sitting fully level this could cause vibe troubles.
With the mains plug, I didn't as such mean the actual socket, but people daisy off to multiplugs, some of these aren't of good quality, a surge protected one is always best imo, but even then some will chain that to another and before you know it you have a power draw coming from lots of different directions, this could cause the extention to break down just enough to send things haywire.
I still think you need to search for some more on the raid config you are using, I can't believe Seagate are shipping bad drives, nor that they have a defect, the one I have is excellent if I'm honest about it, which is why I think it's possible the config is breaking them, or maybe another fault in the pipeline like a MB fault.
10-22-2011 01:25 AM
@fzakbark - agreed.
@cantbecanit - unfortunately it doesn't really matter what you belive, the drive arrived faulty. It certainly could have been damaged in transit but, as with all replacements I have ever received from seagate, It was well packaged so I won't speculate as to how it managed to be defective. I am simply giving you the facts. Additionally the tech I spoke to acknwoledged the fault given the errors Seatools reported and set about handling the replacement immediately. Isn't that why we use these applications... like seatools, to tell use when there are issues with our hardware. ARe you saying seatools is wrong and actually an uneven test bench or faulty power wiring are the cause of the disk LBA errors, and these issues corrected the drive will miraculously start working? Also if the docker had incorrectly reported the drive size or not supported 3TB drives at all don't you think I somehow would have noticed? That being the case would I even have been able to run Seatools at all! Finally please read my posts carefully as I have stated before the raid controller is working perfectly and is NOT causing the issue.
I am not posting here neccessarily looking for solutions as given my situation there really isn't much to be done other than have the drives replaced. I'm here to to report my findings, shed some light for others who may be experiencing similar issues and perhaps give some insight to those attempting to travel a similar path.
10-22-2011 03:11 AM
Now lets not get off on the wrong foot here, all I'm trying to say to you is if they are not being installed with suitable other equipment they will break, it's now 3 out of 5 drives now isn't it?
I saw a thread a few mths back about Raid 5 breaking HDD's and although I can't find it again it was there and that is why I told you of it so you can search for it in your own time because I don't have the time myself.
I myself can only say of my experience, the 3TB I have has been 100%, it was sent to me so I could evaluate it and spot problems, so far as a daily driver for 6 mths iirc it hasn't missed a beat, and also iirc you are the only user whose had any issues other than setting them up.
With the beltsander one, if your drive was level and not on a resonate surface then I'd tend to say Fza nailed it and it's been handled badly on it's way to you and there's not much anyone can do about that.
You also have to appreciate I'm here and you are there, I can't see your rig so I can only suggest, it's like trying to fix a car over the phone tbh, I get limited info from people and most of the time all I have to go on is.....help it's broken and it's a Seagate drive, I just don't want you to end up in an endless loop of returning drives that are going to break down again, now if it's Seagates fault then that's fair enough, but if it's something you are doing to them then it doesn't inspire readers to purchase, and should it be bad operator rather than manufacturing wouldn't you say that is unfair on reputation?
Anyway, fwiw I would simply draw a line under it and refund you if it was my call, sadly it isn't, so the best I can do is to try to highlight pitfalls and hope we get them working properly.
10-22-2011 05:48 AM
Well now, this is why I said 'please read my posts carefully'. Because then you would know that isn't anything I'm doing, my setup is sound and that my controller is not 'Breaking' my drives. I have really been quie explicit in my detail. 3 out of 5 drives failing is a simple fact, and it is totally feasible that the drives have some inherrent issue and are not peforming as specified or (more likely) that I received some from a faulty batch or a batch that were badly handled. This is exactly why I have posted here because this apparently being an isolated incident lends me to belive the later. I have never heard of a controller or array Breaking drives, and though I'm sure it is an interesting read, I imagine that the configuration under those alleged circumstances were not a suggested manufacturers setup, or there were compatibility issues which do not exist here.
The fact that you don't want to believe it is largely irrelevant. This controller is verified by adaptec as compatible with these drives and these were in the drives recommened to me by seagate for my application.
From your very first post you have inclined towards the opinion that somehow my equipment is at fault, or I am doing something wrong, but certainly not that the drives are at fault. Even when I have told you I received a faulty drive out of the box, (which could admittedly be a red herring) you still questioned my setup even with some ridicullous suggestioins about power and uneven benches, after I stated quite clearly that this drive was accepted as faulty by the seagate tech I spoke to after giving him the Seatools error. If your approach is to simply blame the equipment or the user when high failure rates occur why are you even offering advice at all?
10-22-2011 10:28 AM - edited 10-22-2011 10:30 AM
Facts are Jim all you've told me about your equipment is that you use one of these Adaptec 3805 sata/sas host controller card. which I would make my first suspect given you have such a high FAILER rate, real confidence booster that you know what you are doing when you can't spell FAILURE isn't it? this is why I've tried to give you options, options that you simply dismiss out of hand, you've never heard of so it ain't happening is what you are saying, did you know last year Intel bunged out a whole series of MB chips that were duff, took until April this year to get resolved btw.
Now, like I said earlier, if it was my call I'd refund you, it's not my call so I can't, but here's a prediction for you, you will continue to get replacement drives and they will continue to fail, and it'll all be Seagates fault, not Adaptec, not whoever made your secret MB model, just Seagate, because it's convenient to blame the obvious, so here's to keeping the postie busy eh, so like you said, why do I bother to help? it's not like I'm paid to do it is it, I'm just some mug who is ideal to reject because what I told you doesn't fit the plan, and btw, if you knew anything about computing you'd know that good power is the most crucial thing, fluctuating power kills HDD's quicker than dropping them on the floor.
Anyway, I guess I'm done with this one now, you don't need my help, you won't listen to it anyway, and had this been one drive I would agree it's maybe a rogue, but 3, and one standing upright in a dock spinning @ 7200 rpm sounding like a beltsander, oh please!
Hope you get the refund you desire.
10-22-2011 11:32 AM
Ah in the wise words Fry, you know you irked someone when their last bastion of defence is to criticise your spelling instead of your argument.
Now too I see your giving predictions now in addition to your earlier notions of wishful thinking, well I at least I can't fault your consistency. Interesting that you should call yourself a mug though. As for my credentials no comment, but judging from your comments about power I am inclined to belive you are in possession of rather small quantities of information on the subject and probably shouldn't be giving advice on such matters at all.
Actually I only wish you would put the matter to rest. Your last comment shows my assesment of you is correct in that you actually do not believe it is possible to receive 3 bad drives from Seagate, and hence could have saved a lot of trouble with one post and moved on to assist someone else with your knowlege of daisy chained power adaptors and surge protected appliances. Or did you just want to stand on your soapbox and massage your ego with delusions of technical competence?
10-22-2011 04:06 PM - edited 10-22-2011 04:43 PM
With all due respect mate, I'm not the one whose trashed $600 worth of drives am I ?
If you don't know how crucial good stable power is to a PC you don't know much at all, I notice the card is 2007 vintage, a 3TB drive is 2010, that's 3 years and 3 years apart in the world of PC HW is a long time, would you be so good to list the rest of your rig so I can have a dive into that as well, including the Docker
Anyway, I do appologise for my cheap shot at your spelling, but Failer and Docker are not the words I'd expect to read from someone who is obviously as knowledgeable as you are.
Btw, fyi, I was invited to help here because I am helpful and I do know my way around, do you give up free time to help others for nothing? thought not, I just wanted to help you get out of the chain of returning drives, you chose to read me as critical of you, well I'm blind here and you posted limited info even though it asks for specs in my signature, and to sum up, I still think something you are doing is breaking them, the beltsander one was probably because the board in the DOCKER has never heard of a 3TB drive and threw a hissy.
This is from the 6BG/s thread and the guy is spot on, I suggest you have a read, anyway here's Joels comment, I kudos him for it as well, cos I'm generous like that.
10-21-2011 09:45 PM
Mine turned out to be a compatibility problem with the SATA3 enclosure I was using. The drives worked fine with the special Adaptec 3TB firmware and plugged directly into the card. The Seagate drives didn't play nice with the backplane. The vendor said that 3Gb Seagates worked fine with it and 6Gb WDs worked fine. They didn't have any 6Gb Seagates to try but I tried 3TB XTs and 1TB Barracuda 6Gb drives and neither worked. They would all fail building the array and randomly wouldn't even show up."
10-22-2011 04:53 PM - edited 10-22-2011 04:59 PM
Here's some more info for you
This may come as a major shock to all of us who have built their various RAID configurations using standard desktop drives, because standard desktop drives have a tendency to drop out of the RAID array. I talked to engineers and Western Digital, Seagate and Samsung who explained that, when a desktop drives performs internal error correction, the amount of time required exceeds the time that RAID controllers allow before they drop the drive out of the array. When this occurs, the RAID controller writes something (I can't remember what or where) to the drive so that it can't be used again in the array. The Seagate engineer explained that some type of low level reformatting needs to be done before it can be used again.
This is the reason that both Seagate and WD offer RAID drives - they're a $10 to $30 more than their standard drives. Samsung does NOT have a similar drive - I got this from their engineer. These RAID drives limit the amount of time they report to the controller so that they will not be dropped from the array. Two examples of these RAID drives are: Western Digital Caviar RE2 WD5001ABYS, and Seagate Barracuda ES.2 ST3500320NS
Note the "YS" in the Western Digital suffix and the "NS" in the Seagate suffix. Both Seagate and WD also indicate that these drives are built to a higher spec.
Therefore, the person in this thread that indicated that there was no need for special RAID drives was incorrect, probably because non-RAID drives in an array will work.... for a while.
On a personal note, I learned all this AFTER I purchased a pair of Samsung 500GB Desktop drives for my RAID 1 configuration. They've been in my array for the last three months without any problems. That said, I'm going to replace them with RAID-certified drives in the next 60 days.
I just wish that the vendors, like NewEgg, would warn us and educate us with the facts so that we could make informed decsions.
Found here, have a read
And after that please read this through, it'll save you hassle in future
10-22-2011 05:22 PM
10-24-2011 04:31 AM
"Never argue with an idiot, he will drag you down to his level then beat you with experience."
@FZabkar- yeah read all about this before purchasing these drives, figure compatibility reports dealt sufficiently with these issues.
10-24-2011 09:32 AM
Well, this one is a first, unboxed the new drive docked it as soon as it started to spin up could hear the disk rubbing on... what I can only imagine is the a read/write head, unbelievable! Almost.
This one has a cc45 firmware which of course isn't even as new as the last one I recieved so obviously this drive isn't new. I'm kinda starting to wonder if when the drives are returned by others and seagate gets no seatools errors they just ship itout to someone else as a replacment. of course it failed the basic test, not that there was much point running it as you could clearly hear this one was damaged. On its way back AGAIN!
I must be a magnet for faulty drives.
10-24-2011 03:05 PM - edited 10-24-2011 03:09 PM
Ok, I won't argue with you and thanks for the advice because I don't not wish to come down to your level thx.
I'll just remind you that I said you will continue to break drives and have to return them shall I, didn't take long for that prophecy to come true did it?
And now you suggest Seagate send out drives with what sounds like the arm gouging the disk?
I notice you failed to list your rig, I wonder why?
I also notice you thanked Fza for basically backing up what I left for you last time, why no thanks for me mate?
Now c'mon, lets see what you are driving them on, at least then it will give a clearer picture of the situation, you never know if I feel you are rigged up correctly I may even say sorry to you, personally for the third time I would refund you and have done with it, have you raised that possibility?
11-08-2011 10:18 AM - edited 11-08-2011 10:23 AM
Anyhooo... what was it, about two or so weeks since my array failed, changed nothing but it rebuilt overnight - no issues worked like a charm sicnce. I can only imagine I had a bad sector which needed remapping or some such. Seagate appologised profusely for sending me two bad drives, assured me my case is definately the exception rather than the norm, and further verified that the drives sent were brand new and not even certified replacments, so it's a mystery how they both arrived damaged. Spun the drive up while the lady was on the phone so she actually heard it rubbing... was completely dumbfounded it reached me like that, but made no bones about it being faulty. Said she can't do a refund but offered me a constallation ES instead of a Barracuda because of the issues I've had and further that any additional failed drives from this batch (gosh help me) would be replaced with upgrades. Can't ask for more than that.
You know I may have had my share of woes with my drives but the techsup and customer services at seagate is nothing short of excellent. Very very helpful, completely sympathetic and very knowledgable. On the two occasions I've dealt with them on the phone they gone out of their way to resolve my issues.
Constellation ES.2 (edit get my facts right) arrived yesterday spun up without issue in my docker, stuck it in the server rebuilt in the array overnight and so far working like a charm.
I can't say I won't have any more issues though to be fair, I've had none since last glitch and aside from the replaced failures I've had a fairly stable build. At any rate! thank you Seagate not just for givving toss about people that buy your products, but for the excellent after sales sevice and taking this uncanny batch of failures seriously.
Definately restored my confidence.