OK admit it
Who else is tempted to go and drag a dusty box down from the loft... I have an IBM PC with a 10meg drive up there too, and it worked when I last tried it.
Seagate reckons it has found the oldest working Seagate disk drive in the UK: a 28-year-old ST-412 disk drive from 1983. It is the drive for an old IBM PC, which booted up when it was brought down from owner Mitch Hansen's attic in his Ruislip house. The 5.25-inch disk has four platters, eight read and write heads, spins at 3, …
Who else is tempted to go and drag a dusty box down from the loft... I have an IBM PC with a 10meg drive up there too, and it worked when I last tried it.
There's at least one IBM XT up there, plus a TRS-80 with an 8" hard drive.
I have a couple of Macintosh IIci and a Macintosh IIcx lurking in the loft...
Admittedly they're youngsters dating from around 1989, but their disks are SCSII (remember SCSI?)
Said machines intro'd at circa $8,000 22 years ago...
I've got an old Compaq luggable in the loft with a 10mb drive. That worked, last time I powered it up.
I have a dusty box - well, maybe not so dusty - with a Seagate drive from 1996 that still runs fine.
Meanwhile, I have Western Digital drives that fail SMART after three years of use.
In the loft I should have my Atari with it's ACSI (like SCSI but Ataris version) interface.
The problem was that the OS needed around 15 drive letters to address all the space - it had to handle them as floppies if I recall correctly.
Even if I can find it I doubt it would still work though - it needed to boot off a floppy to see the harddrive :-(
Isn't SCSI still alive and well, in SAS (clue: Serial Attached SCSI!), iSCSI (clue: internet SCSI!), and I think used Fibre Channel as well..?
"but disk I/O has become a bottleneck at the platter surface level, and is set to remain that way."
...for a few years until inevitably SSDs become the norm and our children say "your drives moved!?!"
Or maybe even "drives!?!"
maybe seagate can take it back to the lab to figure out the miraculously rare ability this particular unit has demonstrated for a seagate drive
we've had around 75% failure from seagate desktop units under a year old
warranties are all well and good but it's better if you dont have to use it !
Got that right, perhaps the owner of this drive should carry out an experiment by climbing a tall ladder with his disk and dropping it on the Seagate management, just to see if any of them can still recognise reliability if it fell on them.
I have a large pile of dead seagate drives which I won't send back under warranty because I don't want a seagate replacement, I'll buy a Hitachi or a Samsung instead of waste my time replacing bloody seagate drives, the modern ones are all made of cheese. It is a shame Amstrad don't make computers anymore, they would be the natural home for a modern seagate drive.
Are doing the same. Pretty much all modern drives have very little tolerance for failures given the density of the data. It's just too expensive to engineer the same degree of reliability into individual drives which is why the companies are more than happy to supply replacements but never guarantee against data loss. I would be extremely surprised if any modern drive that gets a reasonable amount of use is still working in ten years even five is pushing it. They know that it is far cheaper for customer to have redundant backup drives than a single but very reliable drive and they know about replacement cycles. And then, of course, the market will accept different levels of quality: cheap but maybe not so reliable, expensive and hopefully more reliable from the same manufacturer.
Slagging off individual manufacturers is like slagging off the various mobile networks: heavily skewed by personal experience.
>>"I have a large pile of dead seagate drives which I won't send back under warranty because I don't want a seagate replacement"
Couldn't you get them replaced under warranty anyway, if only to give Seagate some useful (ie expensive) feedback on reliability, and then eBay the replacements they send, or would that involve painful bureaucracy?
Reminds me of the Tommy Cooper joke - "I want to see the Doctor, I said "Doc if I do this with my arm, it hurts!"
He said "Don't do it then!"
There is no way I would trade a Seagate for a Hitachi Deathstar.
Christ, don't buy a Samsung drive.
'There is no way I would trade a Seagate for a Hitachi Deathstar."
To be fair, you'd probably need a time machine to do this. The deathstar era was about 10-12 years ago now. Modern Hitachi drives are fine*, unfortunatly they just got sold to WD.
In my very limited experience current HGST are great (I've got about 10 of the buggers in various stages of aging, some in almost 24/7 use for several years.)
I've had a fair share of Seagate failure over the years too, although my fav drive I ever owned was a Seagate SCSI drive from the mid 90s, all of a gig in capacity, but thin as you like and dead reliable!
infact, I think of all the HD makers I use I favour HGST, most of my current drives are them, I've even got an old 60G GXP from the deathstar era (that had failed in a friends PC) which after a firmware upgrade is still working well 10 years later.
* for a given value of fine, they're as good as any other make anway, and better than some.
Ironically it was Seagate that partly cost Amstrad their reputation for fairly reliable computers.
When Amstrad launched the PC2000 series in the late 80's they used Seagate hard disks. However they were faulty and had to be recalled. That combined with a problem with a new version of MS-DOS caused a lot of bad publicity and they quickly lost their dominant market share.
Amstrad sued Seagate for $100 million and won. But it took something like 6 or 7 years to do that and by that time they pretty much no longer made computers.
" I've even got an old 60G GXP from the deathstar era (that had failed in a friends PC) which after a firmware upgrade is still working well 10 years later."
Some mistake, surely? In 2001 the largest drives were in the 15Gb range - the IBM Deskstars. I had four of them in a RAID0 on an Abit board. No failures, eventually sold the system. You wouldn't have gotten a 60Gb drive until back end 2002/early 2003.
That said, I never had a problem with the deskstars, and believe that the root cause lay elsewhere in the system (iirc we discussed this either on 2cpu or Ars and came to a conclusion that it wasn't the drive per se)
I don't remember if it was a Seagate - the guy I borrowed it from said it came out of a PDP-10, so who knows? The one I had made the PC take about five minutes to even begin to load DOS (1988-ish time frame) because it drew 4 amps from the 12V rail while spinning up, and the PSU was only just big enough to not just shut down from over-current. It also used to drop sectors, so I gave it back and bought a 20MB drive (ST-225) from the lowest cost vendor I could find in the US version of Computer Shopper, located just three miles from home...
Also: arithmetic fail: 10MB per drive divided by eight heads makes 1.25MB per head, not 5.
You know, the scariest part is remembering the model number of a drive I bought over twenty years ago, without looking it up...
More scary is being able to remember the Cylinders, Heads and Sectors for it, without looking it up!
You'd buy a HD and a HD card both bolted to to a steel casing. Slot the whole thing into a PC, this was around 87/88. Sizes where around 8-15MB tops.
Then you'd have to read the manual to find out where the firmware boot address was so you could use DOS to low level format it, before laying a file system format on it. None of this, shove it in and wait for the O/S to ask you about formatting it cobblers!
The drive would run for about 4-5 hours tops then start slowing down as there was no real concept of cooling in PCs back then. So you'd switch off, unplug the HD and use floppies for an hour while the drive cooled down before you'd fire it up again.
I still remember the wonderful metallic clicky sound the heads made as they skittered about over the platters, wonderful noise and great days!
I have one of those too. Unfortunately I no longer have a computer with a full length standard IBM slot so I can't tell if it still works.
I remember the satisfaction of swapping the motherboard's 8086 for an NEC V30 and improving the machine's performance just enough to be able to change the disk's interleave factor so doubling it's throughput.
And I thought 7 years out of a Hitachi Deathstar was good!
The Real Deathstar drives were made when it was part of IBM. I still have a dozen of the infamous drives - all dead and stacked in a filing cabinet.
I don't know whether this is relevant, but I have used IBM and Hitachi Ultrastar (the enterprise version which actually shares much of the mechanical components), and generally they have been quite reliable. I have 3 in an IBM RS6000 44P-170.
See above ^
I had an old ST drive somewhere I used to use as a door stop, I wonder if I can find it... More importantly, I wonder if I can find an ST interface card (and drivers) to test it out!
...when you use a HD as a door stop. Fair play sir, fair play.
Uses hard-drive platters as mug coasters, then reassembles them and uses them in live servers.
and doesn't afraid of anything
>> A real geek?
> Uses hard-drive platters as mug coasters, then reassembles them and uses them in live servers.
...tried the first part of that idea for a bit after my last MythTV related drive failure.
Didn't realize the platters were so shiny...
The novelty soon wore out and I just tossed what was left content to the fact that nearly no one is geeky enough to put the drive back together again (and read my "vital" data).
If someone does bother they will likely be disappointed (Dr Who and Stargate reruns).
... I've got a 10 year old TiVo in which one of the disks is still original Quantum drive (other was replaced when I increased size ... that's a Seagate which is 8-ish years old). Was still working fine when I last used it last month ... though as I've upgraded to the VM-TiVo so its still on but not in use. I'm sure there are several other people with simiar units ... and as these are PVR's the disks have been in constant use over those 10 years and not been spun down.
Meanwhile ... last year bought a NAS to act as a backup drive - disk on that failed after 3 months!
... outlives your TiVo. 16 or 17 year old Mac still working with its Quantum SCSI HDDs, back when Macs were user-serviceable!
My dad complains that he can't get stuff out of that Mac, but the real reason for that is because he threw away all the LocalTalk wiring, including the LocalTalk to LPT adapter we had. FAIL! Though I do wish we had bought an Ethernet adapter back then ... instead we hooked our PC to the LocalTalk network. Oh well...
I had an Intel 80386 (16Mhz) AT machine circa 1985 with a similar 30Mb Seagate SCSI HDD. Having been introduced to fixed disks, (roughly the size and shape as washing machines!), at university some years before, I marveled at how small the drive was, I thought I would never fill that disk with data and I never did. It was bloomin' fast though, well, comparatively. The disk weighed a ton, as did the chassis. They don't make them like that any more, by heck!
Now I'm looking back at that tech the same way I do at a vintage car show - how time files ;-)
I've a Sun E450 power supply as a door stop - does that count for geekiness?
[great 'cos it has a handy carrying handle]
My geek claim to fame is having a failed drive out of my mail server's RAID5 array as a footrest. No, it isn't a very good footrest (it is an old SCSI drive, but not THAT old), but whenever some piece of hardware or another annoys me I can just kick or stomp on the old mail server drive rather than taking it out on a production bit of kit. Though I have found that less issues have been cropping up since I took the drive platter out of a range target (I mean "drive disposed of via nontraditional data wipe methods involving various firearms") and set it on my desk as a warning to the other hardware. Apparently being stepped on daily doesn't send a message, but being ventilated with .45caliber-sized holes does.
"drive disposed of via nontraditional data wipe methods involving various firearms"
sir, I salute you
...the burnt out remains of its predecessor sitting beside it.
I have an old SUN Sparc 20 machine at home which I use as a foot rest. Right size to be comfortable, and weighs a ton relative to its size so it hardly moves on the floor. It was an old broken one someone in the computer department gave me for the purpose when they were ditching old machines.
That said when I started work 11 years ago (as the new PFY graduate I got the crappiest machine in the department) I was given a Sparc 20 as my only computer. It was slow, but it served well as both computer and foot rest. I even kept it in place on the floor as a foot rest when I finally begged someone to give me an Ultra 10 from a batch that were freed up from a load of redundancies.
I still have my original (and presumably still working) Sparc 20 somewhere - as I took it with me when I left. Its probably in my "to-do" pile.
<scurries down to machine room>
You're right! They do have handles! When that servers not used any more I'm having that...
(Anonymous to protect the 'reputation' of this fair establishment.)
The one advantage of the 10 and 20Mb units was the external stepper so when the bleedin' heads stuck to the platters you could physically move head assembly to free the platters without opening the drive! Oh, anyone remember having to rub the spindle eathing strap with a pencil to stop it squealing! Kids these days don't know what a hard disk is! :-)
"Oh, anyone remember having to rub the spindle eathing strap with a pencil to stop it squealing! Kids these days don't know what a hard disk is! :-)"
No. But I remember working the 10 DIP (2 groups, 4 and 6) switches on a 2nd hand one to get my PC to recognise it
At one point I had the 51/4, 31/2, internal and new one all on and the PC recognised *none* of them.
These young people of today blah blah....
what about pulling the sticker off and dropping tiny amounts of sunflower oil from a cocktail stick into the hole to free up the sticking drive just long enough for you to copy everything off?
Now I just feel like an old geek--I maintained CDC equipment and some the CDC 3000 era disk drives were hydraulically actuated. Nothing like regularly taking a 2 foot qtip dipped in alcohol to remove hydraulic fluid from the platters of the refrigerator sized 20mb drive. The next generation used voice coils, ala speakers. Stepping motors were for cheaper printers...
Well, I remember a friend who wanted four MFM drives installed on a PC with an 8bit ISA bus, having to physically break the track on the controller's PCB which signaled for drive 0/1 and connecting it to the drive 2/3 line.
Anyone apart from me old enough to remember drives where you had to park the heads using DOS before moving the machine? My Amstrad PC1512 had one of those in it - last I saw that machine was still working too.
while getting cold hands as you'd just taken it out of the freezer, perchance?
Never knew you could do that.
Seems mileage on Seagates vary. Also on any type of HDD you get nowadays.
Got three Intel servers here, all with Intel RAID. Two have 1Tb HDD's, the third one 250Gb HDD's.
The third Intel server's HDD's have been flaky from the start. The other two is still soldiering on.
We also got a couple of old HP LH3 NetServers - with their original HDD's still performing well after all these years.
My first PC had a 40Mb Seagate - you could kick it, drop it, slap it, it just continue working.
I might be wrong, but it seems that the newer HDD's are more prone to errors and failing - part of the design of using a voice coil instead of a stepper motor. With a voice coil you do need a special platter to keep track information so the drive will know where the head is at any given moment - and once that platter gets corrupted then your drive is also gone.
The higher the aureal densities, the higher the chances of it going south with a bigger load of data at any time.
Anybody remember the infamous Kalok hard drives back then?
the sound they made when you booted up...