• Please review our updated Terms and Rules here

Need a 'Boot and Nuke' program for computers with VERY little memory.

Fire-Flare

Experienced Member
Joined
Dec 17, 2009
Messages
273
Location
Washington State
I work at an electronics recycling facility and we occasionally get a few 'classic' machines in.

We prefer to refurbish working units and send them back into the world, but we don't have the time to dismantle each machine and connect its hard drive to another computer to wipe the former owner's data.

DBAN is great, but hangs on systems that don't have enough memory; Is there another program we can use that will work with as little as 32KB of memory?


Self-booting from 3.5", 5.25", and 8" floppies would be appreciated!
 
I have a Trojan executable that, when run, deletes the first partition on a drive. And, if you then run it again, it will delete what *was* the second partition but since the first partition has already been deleted it sees it as the first partition and... yup, you guessed right -- deletes it. It's under 3K in size so its memory requirement is quite minimal.
 
I have a Trojan executable that, when run, deletes the first partition on a drive. And, if you then run it again, it will delete what *was* the second partition but since the first partition has already been deleted it sees it as the first partition and... yup, you guessed right -- deletes it. It's under 3K in size so its memory requirement is quite minimal.

That sounds promising, how might I get a copy?
 
N.B. Most trojans/viruses can't handle large drives because they don't use the INT 13H extension calls to write data.
So your limit is about 7 GB, which isn't much by today's standards.
 
Plus it probably doesn't wipe the data very well. I don't know how much of a concern that is in this case, though.
 
It's a DOS executable so if that is what you're looking for I can give you a link. It's not a bootable file so you need to be running DOS to use it.
 
N.B. Most trojans/viruses can't handle large drives because they don't use the INT 13H extension calls to write data.
So your limit is about 7 GB, which isn't much by today's standards.

Well it's a start. I have a few machines I've been hanging onto that the trojan can probably get taken care of.
 
N.B. Most trojans/viruses can't handle large drives because they don't use the INT 13H extension calls to write data.
So your limit is about 7 GB, which isn't much by today's standards.
OK, show me a system with ~ 32KB memory and a drive >7GB, or 8" floppies. :)

BTW you have or had this file so you know or knew exactly what it is.
 
For wiping IDE drives, I used to use MAXLLF. Of course it doesn't really LLF all but the earliest drives, but it does neatly zero everything out. It supports CHS and LBA and is a simple small DOS program. I think it is only good up to 127 GB, but anything that large you should be able to use DBAN.

For any earlier MFM/RLL drives, probably the best bet is SpeedStor. Although it probably doesn't recognize every MFM/RLL controller out there.

SCSI systems would be bit of a headache as you would have to find the matching SCSI card drivers. The better ones, however often have a "low level" option built in to a BIOS that will zero the drive similar to MAXLLF.

If the DOS drive happens to be recognized by a DOS boot disk, you can also use Norton Wipedisk. I still use the 4.5 version on floppy disks. That version supports partitions up to 2GB. Later versions support more paranoid wiping methods and probably larger hard drives. I think the 4.5 memory requirements are fairly low, not positive but I would expect 128K or 192K perhaps, and if I recall correctly it works with MS-DOS 2.x.

If you have a system with 8" disks, it probably doesn't have a hard drive. In which case you can simply degauss any floppies if you must.
 
Last edited:
I've used SGATFMT4 from ftp://ftp.seagate.com/techsuppt/seagate_utils/sgatfmt4.zip to wipe a Seagate PATA drive but I don't think it's limited to just Seagate drives. It's only about 79 kB so hopefully it'll run on low memory machines. Here's a bit from the manual:

SGATFMT4 (Seagate Format) is a lo-level formatting utility designed
for AT 286/386/486 systems, only. (If the program is run on an XT,
most likely a stack overflow error message will display.)

SGATFMT4 is designed and LIMITED to work with the following Seagate
disc drive interfaces: ST412 (both MFM and RLL), ESDI (with controller
bios disabled), and ATA/IDE (with certain limitations). SCSI
interface disc drives are not supported. (See the section "ABOUT
DRIVES NOT LISTED")

SGATFMT4 does not use the system BIOS to access the drive, but instead
uses the AT register command set.
 
OK, show me a system with ~ 32KB memory and a drive >7GB, or 8" floppies. :)

BTW you have or had this file so you know or knew exactly what it is.

Yabut-- you need a utility to know the difference and supply the right calls to the API no matter if it's a 16K 5150 or a 16G super-hot box.

cthulhu said:
SGATFMT4 does not use the system BIOS to access the drive, but instead
uses the AT register command set.

And if you have a controller in the PC that doesn't conform to the standard AT I/O port mappings (e.g. SCSI disk) and registers, you're SOL.

If I understand the OP, he wants something that can boot on just about ANY x86 machine and wipe whatever hard disk is in there--and the guy doing the wiping doesn't have to be an expert on what's in the box.
 
If I understand the OP, he wants something that can boot on just about ANY x86 machine and wipe whatever hard disk is in there--and the guy doing the wiping doesn't have to be an expert on what's in the box.

Meaning a program that understands all drive types whether or not they contain their own controller. I don't think such a thing exists. The utility I linked to should work on every type of drive except SCSI and on any PC with a 286 CPU or newer, no BIOS support for the drive required. Know of anything that beats that?
 
My method of formatting drives is something I don't even do in-house, simply because the machine netboots off a server. and I have -no idea at all- how the hell that works. ;)
Now mind you, I always pull drives when I format. Trying to format in machines with unknowns is asking for trouble.
One of my old employers has a junky old pentium III bolted to the wall with a wooden plank hanging out of the case. Inside you had IDE and an AHA-2940UW which handled SCSI. There was no CD drove or even a hard drive. Just cables hanging out of the case for data and power. You netbooted what was essentially a stupid compact linux install with DBAN and it went to work erasing any drive it detected attached to the machine. We later upgraded the machine with a fiber channel card so we could also plug in fiber channel arrays and leave it to nuke all the drives in the chassis at once and eventually a SATA controller as well. The linux kernel took care of the drivers and DBAN just seemed to work without any modifications.
I think the netbooting design is based off the same system used at FreeGeek Vancouver. You might want to contact them and see if they might be a bit more helpful in suggesting how to setup something slightly similar.

For MFM stuff however there was a nasty BabyAT 386 machine with a Western Digital MFM controller with the formatter in ROM. The machine booted off floppy, used debug to enter into the formatter and then we used TH99's drive list to specify disk parameters and let the formatter do the rest. I wouldn't try using formatters that run within DOS itself as then you got an abstraction layer to negotiate that tends to cause weird things with drives that are not yet (and won't be because it's only gonna be plugged in for twenty minutes or so) properly setup in the BIOS. You can leave stuff like interleave optimization and formatting until later if you just need something that will wipe the drive enough to get it OK'd to leave the place.
 
Last edited:
I'd jump for joy since I've never seen an ESDI drive! :)

Really? They're easily mistaken for MFM drives. I think I've still go a couple as well as the controllers. Similarly, how does anyone except a very good tech notice the difference between an XTA and an ATA drive?

And so on. Really, you do have to open the box up. I've run across at least one system where a second installed drive wasn't connected--probably a dupe placed in the box in case the primary failed.
 
I've used SGATFMT4 from ftp://ftp.seagate.com/techsuppt/seagate_utils/sgatfmt4.zip to wipe a Seagate PATA drive but I don't think it's limited to just Seagate drives. It's only about 79 kB so hopefully it'll run on low memory machines. Here's a bit from the manual:

SGATFMT4 (Seagate Format) is a lo-level formatting utility designed
for AT 286/386/486 systems, only. (If the program is run on an XT,
most likely a stack overflow error message will display.)

SGATFMT4 is designed and LIMITED to work with the following Seagate
disc drive interfaces: ST412 (both MFM and RLL), ESDI (with controller
bios disabled), and ATA/IDE (with certain limitations). SCSI
interface disc drives are not supported. (See the section "ABOUT
DRIVES NOT LISTED")

SGATFMT4 does not use the system BIOS to access the drive, but instead
uses the AT register command set.

Thank you, I'm running the program now and it looks good. The ability to verify the drive is very useful too.


Do you have any similar utilities for non-Seagate drives?
 
One of my old employers has a junky old pentium III bolted to the wall with a wooden plank hanging out of the case. Inside you had IDE and an AHA-2940UW which handled SCSI. There was no CD drove or even a hard drive. Just cables hanging out of the case for data and power. You netbooted what was essentially a stupid compact linux install with DBAN and it went to work erasing any drive it detected attached to the machine. We later upgraded the machine with a fiber channel card so we could also plug in fiber channel arrays and leave it to nuke all the drives in the chassis at once and eventually a SATA controller as well. The linux kernel took care of the drivers and DBAN just seemed to work without any modifications.
I think the netbooting design is based off the same system used at FreeGeek Vancouver. You might want to contact them and see if they might be a bit more helpful in suggesting how to setup something slightly similar.

For MFM stuff however there was a nasty BabyAT 386 machine with a Western Digital MFM controller with the formatter in ROM. The machine booted off floppy, used debug to enter into the formatter and then we used TH99's drive list to specify disk parameters and let the formatter do the rest. I wouldn't try using formatters that run within DOS itself as then you got an abstraction layer to negotiate that tends to cause weird things with drives that are not yet (and won't be because it's only gonna be plugged in for twenty minutes or so) properly setup in the BIOS. You can leave stuff like interleave optimization and formatting until later if you just need something that will wipe the drive enough to get it OK'd to leave the place.

I've been thinking about building a system solely for formatting old drives, but I think having an arsenal of programs that can run from DOS might be less time consuming than dismantling each machine to get the drive(s) out.

My boss likes things to be done fast, simply hooking up a vintage machine and running a program is faster than dismantling it to run the formatter on a separate machine.
 
See my post on the previous page about MAXLLF. That woks on all brands of IDE drives.

It seems like the last post on pages never gets seen.

After thinking about it, your best bet for MFM/RLL and SCSI systems is probably just to keep a couple of machines around with well known good interface cards (formatters in BIOS) and long cables that you can just run inside a machine without unmounting the drives. Of course, with MFM and RLL drives, they will then appear completely unformatted to the original controllers.

Edit...shiiii last post again.
 
Back
Top