Anybody have any experience with this? I do a lot of data forensics in my off time (which is dwindling these days, I don't even really have the time to do what I'm asking for information about in this thread, but I'll get to it eventually) but I never have to work with dynamicly sized images outside of VM's since a RAW image is superior in every way for forensics.
Basically what I want to do is make a small raw image file, give it a false size (so its a 1TB file, with a size of only a few MB's on disk) and then format it SquashFS. The purpose of this is compressing my raw disk images in a way that I can still directly interface with them, instead of doing dd into a gz so I end up with image.raw.tar.gz and then having to extract it to work with it, I can dd straight into a raw image file located in the mounted squash image, where I can work directly with the image. This way I can condense the space utilized by my client's images. All of the client images are currently housed inside a 1TB Luks encrypted image, which is about 50% free space itself, though its still 1TB on disk. One of the images is of a 160gb drive, but only 15GB of the drive was actually utilized - I still want the raw image, but I want to compress all those zeroes...
Using SquashFS for the entire array isn't really a viable option because I need high performance from the filesystem for other applications.
Edit: I'm a moron. Sparse files should be able to accomplish what I'm looking for.
Basically what I want to do is make a small raw image file, give it a false size (so its a 1TB file, with a size of only a few MB's on disk) and then format it SquashFS. The purpose of this is compressing my raw disk images in a way that I can still directly interface with them, instead of doing dd into a gz so I end up with image.raw.tar.gz and then having to extract it to work with it, I can dd straight into a raw image file located in the mounted squash image, where I can work directly with the image. This way I can condense the space utilized by my client's images. All of the client images are currently housed inside a 1TB Luks encrypted image, which is about 50% free space itself, though its still 1TB on disk. One of the images is of a 160gb drive, but only 15GB of the drive was actually utilized - I still want the raw image, but I want to compress all those zeroes...
Using SquashFS for the entire array isn't really a viable option because I need high performance from the filesystem for other applications.
Edit: I'm a moron. Sparse files should be able to accomplish what I'm looking for.