CVS keeping photos for years....

[odysseus]

Oh, I remember when degaussing was, and in some places still is for good reason, the main wipe standard.

Regarding SSD drives and wipe, this is why built in passive encryption is becoming more prevalent for them. However that might not be robust to some, so people for secure data that is portable should consider proper encryption software, and certainly for those where encryption is not passively a part of the SSD drive.

In fact if you have secure data on your workstation you need to protect, encryption is a far better strategy since it protects is against loss or theft, instead of only when you decide to toss it out and wipe it.

Regarding wipes, I believe no one has publicly claimed recovery of a drive after even a one pass random bit wipe. Sure, perhaps serious labs in a few highly sourced areas might theoretically, but that is another conversation.

[Wherryj]

Back in the late 90's Eccard Drugs (before CVS bought them) Had a best of the best photo album. The photo guys were saving copies of the nekkid lady prints in a special photo album. Some ex employee ratted them out to the local media. Unfortunately, although the technology has changed, people wont change.

I managed a photo lab in the 80s. I wouldn't allow this to happen, but I know that it probably still did-just without my knowledge. As you say, there are always people like this.

[darkgrue]

Solid State Media does NOT retain information except in two ways: One- the file table and Two- Unwritten to memory.

All you need to do to wipe solid-state media is flip the bits. Unlike magnetic media it does not retain "ghost" images of files in any kind of magnetic field.

Hard Drives can and DO, through many wipes in fact. If data protection is of concern use solid state media because if you want to destroy it all you need to do is a bit flip and the data is gone, FOREVER. No forensics in the world has the quantum technology needed to harvest data off a table wiped and bit-flipped solid state memory device.

Not entirely true. It's true that solid-state media doesn't retain information like the hysteresis loop present in magnetic flux density. However, the devices can retain data for other reasons, and certain failure modes are also significant.

Solid-state media doesn't have the same sort of data reminance issues that magnetic media does, but they do exist.

Solid-state devices have abstraction layers (the most common of which is wear-leveling and bad block management) that stand between the user and the actual physical memory cells. Standard erasure methods only write over a file location, while the SSD wear-leveling technology may specify that a write operation (of all zero's, or even a pattern) is written somewhere else entirely. Multiple passes may get written to yet another location, and so on. The data, unbeknownst to the end user, is still there and able to be recovered by those with the right resources.

Similiarly, there are several different failure modes with bad blocks on solid-state devices. And generally, sections are marked bad by block, not by individual bad bit, so even though a section of memory has an error (bits stuck high, stuck low, etc.), there could be a good portion of the block that can be read. Block sizes vary, but can be pretty significant, into significant numbers of KB - a bad block can contain meaningful readable data. The bad block mapping will normally prevent formatting or other commands from accessing those marked blocks - again, unless you have the right tools.

Sometimes it's easier to recover data, than it is to sanitize it.

Some commercial devices are getting better - and allowing things like secure erase (using a special utility function, not a normal format command), which does override some or all of the management functions for a true erase. Mostly this is being seen on newer SSD, but there are still many devices out there (especially memory cards and thumb drives) that have no provision at all for that (or that don't work as expected).

On the magnetic side of the coin, any utility that claims to be DoD-compliant or offers a "NISPOM" or DoD 5220.22-M erasure method... they're citing a very old and superceded standard. It's probably just fine for civilian use, but there is currently no provision for sanitization [resulting in reuse] of magnetic media in the classified arena. It gets destroyed (by incineration or disintegration) or gets put in a suitable safe for all time.

[JDay]

Some commercial devices are getting better - and allowing things like secure erase (using a special utility function, not a normal format command), which does override some or all of the management functions for a true erase. Mostly this is being seen on newer SSD, but there are still many devices out there (especially memory cards and thumb drives) that have no provision at all for that (or that don't work as expected).

As far as I know pretty much all ATA devices support the ATA Secure Erase (it is part of the specification for ATA) function. This includes SSD drives, most of which are hardware encrypted. When you secure erase a hardware encrypted SSD drive the drive will erase the current encryption key and generate a new one. All Sandforce based drives work this way. Intel, Crucial and Samsung SSD drives are the same afaik. You only need to make a bootable Parted Magic CD to do this.

http://www.zdnet.com/blog/storage/how-to-really-erase-a-hard-drive/129

[meaty-btz]

Not entirely true. It's true that solid-state media doesn't retain information like the hysteresis loop present in magnetic flux density. However, the devices can retain data for other reasons, and certain failure modes are also significant.

Solid-state media doesn't have the same sort of data reminance issues that magnetic media does, but they do exist.

Solid-state devices have abstraction layers (the most common of which is wear-leveling and bad block management) that stand between the user and the actual physical memory cells. Standard erasure methods only write over a file location, while the SSD wear-leveling technology may specify that a write operation (of all zero's, or even a pattern) is written somewhere else entirely. Multiple passes may get written to yet another location, and so on. The data, unbeknownst to the end user, is still there and able to be recovered by those with the right resources.

Similiarly, there are several different failure modes with bad blocks on solid-state devices. And generally, sections are marked bad by block, not by individual bad bit, so even though a section of memory has an error (bits stuck high, stuck low, etc.), there could be a good portion of the block that can be read. Block sizes vary, but can be pretty significant, into significant numbers of KB - a bad block can contain meaningful readable data. The bad block mapping will normally prevent formatting or other commands from accessing those marked blocks - again, unless you have the right tools.

Sometimes it's easier to recover data, than it is to sanitize it.

Some commercial devices are getting better - and allowing things like secure erase (using a special utility function, not a normal format command), which does override some or all of the management functions for a true erase. Mostly this is being seen on newer SSD, but there are still many devices out there (especially memory cards and thumb drives) that have no provision at all for that (or that don't work as expected).

On the magnetic side of the coin, any utility that claims to be DoD-compliant or offers a "NISPOM" or DoD 5220.22-M erasure method... they're citing a very old and superceded standard. It's probably just fine for civilian use, but there is currently no provision for sanitization [resulting in reuse] of magnetic media in the classified arena. It gets destroyed (by incineration or disintegration) or gets put in a suitable safe for all time.

Indeed, I was cutting corners in my description. Biggest issue is the bad-block data maintaining.

I still warn people if a drive fails, smash the chip or people can recover data from it. Unlike most people I do have the right tools. I often forget that most people do not. Actually, there have been times I would have loved to recover data off a flash device but, of course, was unable to.