Wipeing drives (zero-fill) takes forever on 1TB platter drives on USB 2.0...

[the86d]

Got approved for a TX-Facility transfer, so I am wiping drives that have been sitting in the computer room... for eWaste.

OMFlippingGolly... ... ...

[Woodymyster]

They still allow us to drill the platter 3 times. Just a tad faster...

[mrkubota]

https://www.youtube.com/watch?v=PS3sAWF5S1I

[Dr Pete]

Ever try a BIG magnetic chuck?

Quick and painless, not quite as fun as punching a couple holes thou

[mrkubota]

we used to have big magnetic tape degausing tools to bulk erase data tapes. Don' t know if they'd work in a hard drive though...

[SnipTheDog]

They still allow us to drill the platter 3 times. Just a tad faster...

A metal bandsaw might be faster yet.

[dizzyblonde]

A degausser and drive destroyer press would be faster.

[the86d]

Most modern drive platters I have disassembled are glass-like, with a sprayed on rust coating. Just a punch is required to shatter.

I am just running Slackware in VMware player, and I can just minimize it, and don't have to dedicate a computer to the wipes.

Still going...

[Kyle1886]

Since I get collect the magnets out of old drives I do the gorilla method---brute force destruction.

Yes, a bit slow but there is a little satisfaction of breaking something I don't need to fix and get something useful in return.

Kyle

[gwanghoops]

disassembly of mechanical devices is always fun

[the86d]

What is weird is that the 1TB drive ("newer" drive) is pushing only 11.3MB/Sec, while an OLD beater 40GB 2.5" laptop drive (out of a single-core Gateway Laptop, I think it was) is chooching along at a sustained 20.1MB/Sec, and both are on the same model USB 2.0 "JMicron / JMicron USA USB to ATA/ATAPI bridge" devices, on the same powered USB 3.0 HUB...

I DID set the "dd if=/dev/zero of=/dev/sd..." block size of the 1TB drive to 4MB, and the 40GB to 1MB, but in my testing this usually doesn't matter much.

Because these are connected via USB I can't check the cache sizes, via CLI.

[DaveInOroValley]

Drop in a smelter?

[lordmorgul]

[Robotron2k84]

You are better using /dev/urandom than /dev/zero on such an opteration for two reasons:

1) /dev/urandom is only about 20% slower than /dev/zero on a 500 MB/s SSD, which would allow you to keep maximum speed in spinning rust, while offering pseudo-random data to write instead of simply zeros, because…

2) If you understand magnetic domain recording, you would know that writing a single value, once, over an entire drive allows the complete contents to be resurrected with relative ease. Either multiple passes, or random data are required to impact enough of the magnetic domain to push the average gauss values solidly into one bit or the other. Otherwise, the prior states can be determined, solely by statistical analysis.

Another win for SSDs: not only will zero-filling erase all contents permanently, but all I need is a hammer, once, to the logic board, for permanent destruction.

Your cheapest and best bet for destroying platter data is renting an industrial shredder.

[Woodymyster]

I dont know the OP's situation, I was just referring to state and federal guidelines dictate how a hard drive is to be decommissioned. They still allow us to put three holes in the platter and deliver the carcass to the DLA at the Marine base in Yermo. You can degauss or do random info, but they have to be done a multiple cycles and take forever.

As far as using a bandsaw, no, three holes is the fastest as most HDs have voids, multiple metals, and different thicknesses that destroy bandsaw blades.

If I had to completely disassemble a HD, I've been able to in 3 minutes without destroying anything, just using torx drivers and a press for the press fit motor types. As far as punching or drilling, usually under 20 secs.