Has a Seagate Middle East sales guy just let slip that 5TB drives will be here in three months? He's identified on a video uploaded to YouTube as "Christian from Seagate", and his video interview has been mentioned on Twitter. The Tbreak video, below, records an interview at GITEX '11, a technology event in Dubai. Seagate has …
If they're coming from Seagate...
...I don't bloody want them. Just watched a Seagate GoFlex Home 2TB network drive (that never worked well at the best of times) just die completely after 10 months of relatively light usage (relatively light because it was so damned useless in the first place).
Without a doubt the most shoddy, damned awful piece of shite product that it's ever been my misfortune to own. If anyone is thinking of buying one, just buy a bunch of A4 notepads and a lot of pencils - you'll be better off. If that's the best that Seagate can do, then any future 5TB drives will probably be best employed as door stops.
yes, i agree, my bad experiences with HD's seem to be with Seagate's. I find WD Green Power drives great for home use (storage), obviously not performance drive, but cool and quiet and perfect for streaming media.
SSDs for performance now anyway isn't it?!
If you'd bought 10 of them & had a decent proportion fail on you then fair enough, but just getting unlucky with a single drive & whinging like a 6 year old with a skinned knee makes you sound like a bit of a tit.
My fiery icon's bigger than yours.
Drives from *any* manufacturer can be prone to failure, I've had Western Digital & Hitachi go tits up after a matter of months - it's just bad luck.
On the other hand, I've still got a 1gb seagate IDE drive that works, despite the fact I used it as a portable drive and would sling it in my bag to take data back and forth from work/friends - this is 10 years back, way before usb sticks with that capacity. I dropped the damn thing many times, just kept on going.
I recall the 2.1gb Seagate Baraccuda drives from many years ago - early 90's I think?
They were virtually indestructible - much sought after!
Normally I'd pretty much agree with you. In fact, with hindsight, I do indeed feel like a bit of a tit for buying that useless pile of junk without checking it out in more detail in the first place.
Having said that, just look at some of the other comments here or take a quick skim through the fairly large number of reviews out there on the 'net complaining about the increasingly poor quality of Seagate's output in recent years. I've pretty much lost count of the number of internal (IDE, EIDE, SCSI, SATA, whatever) and external (USB, Firewire, NAS) disks that I've bought over the years from various manufacturers and, you know what, the only two that I've ever had bother with were from - you guessed it - Seagate. The first one (a 500GB USB drive) I just put down to the normal statistical run of bad luck. However, the GoFlex Home unit that I bitched about at the top of the page was, undoubtedly, the poorest such product I've ever made the mistake of spending money on. Not a mistake I intend to repeat and one that I will happily continue to recount as a cautionary tale to others - regardless of whatever the merry hell you or anyone else thinks.
Of course, YMMV and caveat emptor, etc. etc. In my case, the caveat involves avoiding most of Seagate's products in future.
Sounds like typical salesman speak.... i.e. a load of bollocks.
Don't trust them
That's just too much data too loose at once. Seagate drives are horrendous of late. It would be too clostly to purchase a 5TB drive from Seagate and a spare while waiting for the inevitable click of death that will be along soon. They should work on first making a reliable drive, thn making a larger drive. What the hell happened to them anyhow, their drives used to be great.
You're forgetting The Rules.
Rule number one of data storage: If you don't have your data in at least TWO places then you don't have your data.
If you stick everything that's precious on a disk of *any* size made by *any* manufacturer and don't back it up then you deserve all that you (don't) get when the disk fails. All disks will fail eventually.
I'm not being holier than thou either - it has happened to me, cost me a lot of money and time, but I've learned my lesson.
I thought the rule was "if your data isn't in three places then it doesn't exist"
Disks of such size are a liability
The experience on my server farms shows that Seagates have less of a tendency toward failure than Samsungs. But unlike Samsung and WD disks at least Seagates they don't line about their health:
But the main point is that multi-TB disks are too much of a liability, even in RAID arrays. With the multi-day array parity rebuild times, the probability of additional failures during the rebuild increases to unacceptable levels.
I thought the rule was "if your data isn't in two places, then the quantum entanglement must have decohered".
If your data isn't in n+1 places then it doesn't exist.
That's why google gives yours to everyone.
yeah... it is three
one copy of your data on your local machine plus a copy on a local backup device or server then a third copy off site !!
Maybe Maxtor was what happened to Seagate?
"What the hell happened to them anyhow, their drives used to be great."
All drive companies seem to go phases of being good and bad. A few years back, Seagate seemed to be pretty good.
Perhaps it's not a coincidence that the decline seemed to start around the time they took over Maxtor a few years back. One of my first thoughts was that this would be a problem for Seagate's reputation if they started selling drives made in former-Maxtor facilities under the Seagate brand, as I thought they would. (Rebadging could be done overnight, bringing that factory's output up to scratch would take longer).
After 5 years I'd have thought that they'd have been comfortably absorbed, but maybe Maxtor tainted Seagate?
My Dad had some Seagate drives a couple of years back- I'd rather have gone for WD or something else, but there was limited choice in IDE- and had two or three fail sooner rather than later. The WDs he replaced them with are still fine.
I first used Seagate drives when the 20mb MFM was the standard. They never admitted the 'stiction' problem, where you had to tap the drive to get it to start spinning, even when NEW.
Next I had a 120 MB Seagate drive that lasted two months, then Seagate charged me $25 to replace it.
Seagate drives were always weak, especially compared to Western Digital's offerings. I stick with WD for the foreseeable future.
And One Of...
...the three places should be in another city.
@tomn8r "That's just too much data too loose at once."
What does the word loose mean, in this context?
To shoot, discharge, let fly.
It's what happens when you open the case and spin up the drive. Centifugal force and all that. Bits go all over the place. 40 trillion of them are almost impossible to clean up.
That's how I read that sentence as well
"That's too much data, too loose, at once."
I pictured all these tiny sparkling bits, a la Tron's "Bit" but much smaller, all scampering away beeping and bouncing from a newly opened hard drive. Maybe if they hadn't been so loose, if they'd perhaps tightened the bit restraints a little more, then you wouldn't have had so many of them escape!
definitely _not_ in the building next door...
like they were doing in the World Trade Center...
That's why you use RAID61
If you had 20, 5TB drives, you would have to loose 4 drives at the same time to loose data under RAID61.
So 20x5 = 100TB - 20TB for parity overhead from the RAID6 divided by 2 for the RAID1= 40TB of RAID61
Disk drives are a mixed bag, there's always going to be failures no matter who you go with. I've been lucky in that my last failed drive was an IBM 75GXP Deskstar AKA Deathstar back in 2002/3.
Since then i've owned a *load* of drives and have not had one single failure.
Seagate actually won my respect big time a few years back because at a time when vendors were cutting their warranty Seagate upped it to 5 years for 3.5" drives. That's putting your money where your mouth is.
This is the thing, ask any body about a brand of drive and you will always get a group of people saying the won't touch them and they're the worst drives going.
I've had drives fail on me from all manufactures. Seagate, WD, IBM/Hitachi, Maxtor (In their day) and Samsung all make decent drives, some of which fail. Failing drives is the nature of having a lump of metal spinning at 7500 RPM. I think the only drives I've never had fail are Toshiba ones but I've dealt with far fewer of those over the years so I don't think they're any less likely to suffer an issue.
Once you start dealing with drives in hundreds rather than 2 or 3 you'll see that they are all very similar.
I own two SSDs one of them failed inside the first 3 months, the other one bought around the same time and the replacement for the first are fine.
Doesn't matter how big your drive is, if you want to keep your data you need to have it backed up. If you can't afford to be without your data whilst a broken drive is replaced/repaired you also need to be running RAID.
I dealt with drives by the THOUSANDS, there is a much higher failure rate on Seagate drives, much higher.
Ahh the deathstars
Me, my cousin, and a guy I knew back then bought the 40 gig ones when they came out. We all got 2 each
My cousins majority of the sectors went bad within 1 year on one of this drives, other failed completely both were warrantied. The 2 replacements same stuff bad sectors popping up left and right after he lost some school work he flung em out.
Mine (I still use 1) the one had lost the majority of its space due to bad sectors, other (the one I still use) has a few bad sectors still. I've low level formatted them 8 times each cause that oddly repairs the bad sectors. The one last time I used the tool to repair them failed the repair(why its no longer used).
I was using the 2 drives for over 6 years in my network server which has only been down for mother board upgrades(2 times for 5 hours). I also have a WD drive, and segate drive running the same amount of time, and are used way more.
Other guy I knew with them ripped them out when he got his first few bad sectors, and sold them on ebay after running the drive repair program.
But agree on the rest I've seen various drives of all styles crap out or have issues. IBM's just had the highest rate I ever saw period with such a small sampling(100% out of 8 is pretty bad). Seagates I've got 3 in my current PC (2X1TB, and 1X2TB), and not had a single issue with them so I can't complain(now watch one die the second I hit submit).
You'll never fill that!
That was what an electronics engineer said to me when we took delivery of our first PC with a hard drive, a massive 10 MB...
And we never did!
That is, at home we started with a 5 MB HDD connected to the dual-drive IBM PC. We never managed to fill that disk, and couldn't see how it would even be possible.
Later, we upgraded to an IBM PC XT with a 10 MB HDD, and then we filled it, of course. :)
My first PC (10Mhz 8086) had a 32 Mb hard drive, and at the time I was programming using borland & mickeysoft C, not that I needed the 32mb as both compilers, with built-in editors and debuggers*, lived quite happily on a 5 1/4" floppy.
Within 3-4 years mickeysoft were up version 6.0 of C and a 40mb install.
All on 3.2/3 DOS
*I think the mickeysoft debugger was codeview and may have been a separate application
> "It is hardly a revelation that hard drive capacities will increase in the future or that sometime soon there will be a 5TB offering."
units, you nit!
is it just me, or does he seem to confuse megabytes and gigbytes quite a bit?
I want enterprise class 5TB drives for use in RAID arrays, I wonder how long we'll have to wait for this.
They have the technology now so they should release them ASAP. I'm finding it more expensive to keep using bunches of crappy 2 TB drives in limited size 1U servers. 5TB drives are required immediately, especially for offsite backup servers.
Once these new drives do come along I wonder if they will be supported by the existing RAID cards.
How long are you projecting a rebuild to take on 5TB disks?
Their quality went down the pan when they replaced all the techies with bean counters and people with business degrees
I could use a RAID 0 with two or more of these 5TB beasts....
... to record 5 minutes of a game in FRAPS in FULL HD glory.
If you ever used FRAPS, you know what I am talking about.
Hint: don't forget it RECORDING, or anything short of server storage capacity (several TB worth of it) will fill up in a matter of 1 hour or less. 30s will grant you a cool 1GB raw .avi movie.
And that's just for a scratch disk. Then you must encode it in some sensible form like H.264 or wmv or whatever.
Used to be ...
Larger disks were more reliable.
We passed that point a while back. I think that to get 5Tb disks to market quickly will inevitably result in corners being cut and compromises being made. Net result: huge capacity drives that last 6 months.
Seagate drives do seem to have a failure rate higher than either Hitachi or WD. Don't quite recall when that started but it seems to have been objectively and subjectively true for a number of years, now. I may be wrong.
I agree, but I was blaming it on the various sleep modes..
So... I never let my own drives spin down, ever, and only reboot for updates every 3 months, and try to never fully power down (battery backup power helps), and I don't change the orientation of the drives after they have been formated .
Since doing all those things, I have only lost 1 drive out of 12 in 6 years. My drives have been 4-500GB WD, 4-1TB WD, 4-2TB Seagates all running software based RAID5.
One of the first things I do is switch off power save on the disk on a Windows install. I have no hard evidence, but the sound of the disks, which appear to literally get their power cut, is enough for me to have issues with it.
My Experience is that Seagate are usually very reliable.
Seagate's main weakness seems to be triggered by an unstable power supply coming into the property (for residential users), ie Card meters. I have only seen three Seagate Disks fail in 10 years, and all of them were subject to this environment, and all failed within 12 months of being in such an environment after running with no problem before this.
Its the only case where I can say Seagate may be inferior and might not be a good choice.
I've just pulled out the spec and pricing for a system that we installed for a customer in 2001 - 5.3 TB for the princely sum of £293,000 (+ VAT).
Ok, so it was RAID-5 and had FC fabric to three servers, but even so.
I'll get my coat.
Personally I only buy Seagate drives, after having issues with Western Digital drives over the years. And I've never had a Seagate drive fail, I've always retired them through obsolescence. However, my brother has had exactly the opposite experience and will only ever buy Western Digital. It really does come down to personal experience.
My latest Seagate is a 3TB, which sits alongside my 2TB, 1.5TB, 1TB and .75TB drives. Meanwhile, I've had quite a lot of problems with SSD drives from OCZ. I have a Vertex 3 and get bluescreens, usually every couple of days - my two previous Vertex LE drives both failed because of firmware (expensive failure at over £300). The speed is awesome and I always backup any important data via Dropbox, so I don't take risks. Still, SSDs have a long way to go til I trust them.
5 year warranty
These Constellation-series drives come with a 5 year warranty. That suggests to me Seagate are pretty confident in their reliability.
I've had 16 2TB Constellations running in two NASs now for over 4 months and no issues yet... only time will tell. I'm not going to lose sleep over one or two failing either, since they're in RAID6+hot spare
>30s will grant you a cool 1GB raw .avi movie.
That's a crock of shit. Speaking as someone who's part of a fansub group, we use 2GB transport streams which must be encoded into things such as Hi10P, x264, XviD, etc.
2GB for about 24m, maybe a little more with the commercials.
If you're getting 1GB for 30s then you're doing something wrong.
write speed the issue for rebuild
Why can't a 5-platter disk be 5 times faster than a single plater one? 10 heads, concurrent writes, fast rebuild. No?