Strange request...

[BlackST]

Goal : make a BIG batch of drives unusable and data difficult or close to impossible to recover , satisfying these conditions :

- physical destruction NOT AN OPTION drive must be integer and not disposed
- no electrical damage allowed as...
- ...further refurbish should be possible later but it is not important at the moment
- erase not an option , takes too long.
- should be difficult and unlike to access data by platter swap
- should not be possible to easily access data with low cost tools OR playing with ROM
- each drive should be processed in max 2 minutes, the quicker the better.

Drives in use :

Toshiba GSX 2,5"
Momentus 5400.6
Hitachi CLA
MAXTOR D740X-6L 3,5" (don't laugh!)

In the beginning I thought it could be relatively easy, effectively it seems a bit more challenging than expected...

[jono-ats]

Suppose you were to back up the modules related to the translators, and replace them with those from a different drive?

That would make it very hard for most firms to recover them and you would have the capability to fix them later.

[BlackST]

Thought about it, but the risk is to have data still accessible albeit scrambled and translator is easily regenerable. Backing up fw takes also long on the momentus.

The easier seems the 740...

[BlackST]

Update ... for the 740 and the Momentus found a solution that seems the "Perfect PITA" for DR... Drive calibrates, id correctly and have legit and error free SA and ROM, just no access to data ( UNC on every imager or software, no way to get data even tinkering with ECC but servo is unaffected) and it takes very little time each... now time for Toshiba and Hitachi.

P.S. it was also not possible to keep backup of every FW due to the large quantity of drives.

Hope they will not come back for refurb all together

[ShaneWard]

Take the PCB boards of the drive and store in a separate place. Take the main drives and store them inside a bank volt for future use. When you want to use the drives again just put the PCB boards back together and then wipe the drive by: EG the easiest way would be to copy a load of movie files to the drive until its full. then delete them all.

[MindMergepk]

Goal : make a BIG batch of drives unusable and data difficult or close to impossible to recover , satisfying these conditions :

degaussing

[DR-Kiev]

Goal : make a BIG batch of drives unusable and data difficult or close to impossible to recover , satisfying these conditions :

degaussing

Thats too simple, the task was with option ...further refurbishing.

[HaQue]

Take the PCB boards of the drive and store in a separate place. Take the main drives and store them inside a bank volt for future use. When you want to use the drives again just put the PCB boards back together and then wipe the drive by: EG the easiest way would be to copy a load of movie files to the drive until its full. then delete them all.

doesn't satisfy "- each drive should be processed in max 2 minutes, the quicker the better."

If you do find a solution, would be good to hear what you found.

[Tomset]

http://www.epos.ua/view.php/en/products_epos_lavina

[BlackST]

No degauss, it would have destroyed servo too.
The drives will be recalled for refurb later.

Anyway I found a solution that proven to be effective (the samples sent for checking were absolutely unreadable despite the fact they were full functional , ID and calibrate, the customer said they had to give up after trying with both Sw and Hw tools and playing with FW...).

The time to process each drive is about 1 minute or less for the 740, about 1 minute and 45 seconds on the other, add 10-15 SEC. each for connect and disconnect.

@ShaneWard
I could not keep pcbs, the drives have to be sentback integer.

[deftrue]

You may erase ROM if there are adaptives in it (keep copy for refurbishing), or erase one single module of SA adaptives if not. Is that an option? As long as it is unique data essential for user data access and you keep copy, that could be a quick option.

[BlackST]

Copying Rom imply a lot of time, erasing means inoperational drive, my method give out operational but unreadable drive instead .
Also, I did not want to keep copies of firmware.
At the moment I am working on the next step...

[deftrue]

How was it done, finally?

I can think of a lot of ways, but not sure how quick they are right now... self-test should be able to restore functionality but not data, AFAIK.

[BlackST]

in fact refurb, later, must restore functionality not data, that is not important right from start. In a nutshell, the 'Appliance' has an hdd with data inside that must be rendered inaccessible/invalid prior to shipping. Later, if they don't want to replace all the drives, they have to be all refurbished. The problem would have been the time to erase all drives or replace all drives , prior to shipment. Don't ask me why or what, I just tie the donkey where the owner say to... .
Just told them that fixing and rechecking dozens of disks would not be instant job ... crossing fingers...

[Doomer]

Changing boot adaptives and/or SA location in NVRAM for Hitachi should be enough to make data and SA inaccessible but further refurbishing easily achievable. Although experienced DR pro with enough motivation could still get to data.
There is a way to make only UA unreadable but that makes data recovery from such a drive much easier.

Toshiba can be killed by modifying adaptives in ROM, considering that there are no good pro tools for those drive, data is likely will be inaccessible even for experienced DR pro but still with enough motivation, some bits of data could be retrieved, IMHO.

[northwind]

For Seagates, just take off all pcb's from them, scramble them, and then screw them back randomly

[BlackST]

Too much time... Made it in very little time and all the devices are shipped. Now waiting for call back to make all them work again : it's just like being hired to dig a hole, wait and then being hired to fill it again

Btw : got a call "hey, it is not so secure... if I erase by zerofill the LBA the surface is accessible again..." "INdeed... didn't you ask for data being inaccessible?"...

[fzabkar]

For Seagates, just take off all pcb's from them, scramble them, and then screw them back randomly :mrgreen:

Just match the serial number in flash memory against the serial number on the label.

[BlackST]

CASE WAS ALREADY CLOSED.

Time to screw and unscrew a pcb = 1 Minute

Time to perform the "BlackMAGIC" = about 20 to 30 seconds + 2 to connect/disconnect cabling.

Why hassle ?

(P.S. 2 CABLE SETs worn out during the job...)