Platter Crash on an IBM Death Star HD (2003) | Hacker News Reader

Platter Crash on an IBM Death Star HD (2003)

astro.ufl.edu

37 pointsluu24 comments

People would pay good money for that kind of secure erasure if it could be triggered on demand. :)

... and not by malware :-)

Definitely something you want to trigger by physical contact, not pure software.

There was a company which used to make hard drives for the military (perhaps they still do, but nothing on Google) which had a physical pull tab on the back which released an acid into the drive enclosure. Anyone remember this? I read about it perhaps 10 years ago.

Sounds vaguely familiar to me too.

Was it an acid? I've heard of thermite or magnesium. Same idea though

I assume was something like acid since it released a liquid into the cavity.

I remember tearing into one of these that died, and found similar things. What was interesting too was that it seemed to be caused by a screw coming loose. The entire head assembly rides on a bearing, which is bolted to the chassis, usually on both ends. If I remember right, this one was only bolted on one end, the screw didn't have any thread locking compound on it. It came loose, instant head crash.

Back before I could afford the minor increase in cost of ECC ram, and the larger cost of using RAID 1 or 5 or 6; I got bitten by this very issue in my personal hardware.

Just one of many times that I've been taught. Any data that is actually /valued/ should have a backup. Redundancy in operation is more a hedge against hardware failure, and unplanned downtime.

Integrity costs; pay for it. This is true everywhere.

It's not really a minor increase in cost for ECC ram; the CPU and motherboards that support ECC ram are considerably more expensive. To top it off, many times motherboards cannot support the full capacity of ram with ECC installed; my workstation can accept up to 8GB DIMMs with non-ECC and 4GB DIMMS with ECC.

What was that blue stuff in the drive?

Phlogiston.

I believe that it would be lubricant from the hydrodynamic bearing[1] inside the hard drive.

[1: https://en.wikipedia.org/wiki/Fluid_bearing ]

Thanks, that was interesting to read. I take it that helium-filled hard drives work in a similar way?

Exact same way. The only difference is that helium is much less dense than air, so there's a lot less drag and thus the drive is more efficient.

That's just glue to hold the shielding plate on. You'll notice its nowhere near the bearings; in fact, it's on the outside of the case.

So I guess the disk platters used a glass substrate, coated in a layer of magnetic medium?

Indeed so.

http://www.computerhistory.org/groups/storagesig/media/docs/...

'The drive, a 7,200RPM Deskstar 75GB drive, was released on 15 March last year [2000]. At the time, the press release announced excitedly that the drive was "the first IBM drive to use glass disk platters instead of aluminium ... allowing the recording head to read smaller bits of information that are packed more closely together. In addition, glass disks are more stable at higher speeds".'

> was asked to see if we could determine the cause of the crash, as well as to determine if it would be possible to recover any of the data on the drive.

If it had been possible, it would have ceased being so after you took it apart at your desk!

Untrue. I once opened up a hard disk, left it open for a day, closed it up again and used it without errors for another year (though never for important data). And no, not this was not inside a cleanroom.

Ahh, I remember this scandal. What's funny is that there were tons of people affected by this from what i recall, because these drives were initially billed as some of the best hardware money could buy.

I actually had one of these disks, and used it as my main system drive for 6 years before finally retiring the entire system.

Yup some of them were fine. It was only certain capacity disks that had the issue. They solved the primary problem eventually. By then the reputation was already tarnished. "Death Star".

I wonder how many bits each particle of dust represented?

Reply on news.ycombinator.com