A number news storys have pointed out how easy it is to have your data compromised via a stolen laptop. After about a dozen of these lost laptop/missing goverment data stories, I got to thinking about my own data. I carry my laptop everywhere and it contains a lot stuff that I wouldn't want an identity thief to get his hands on. Granted I run linux and my simple login password is probably enough protection against 99% of laptop thieves, here is how I decided to protect my data.

First I installed truecrypt. It is a filesystem encryption software that runs on both linux and windows. Truecrypt works by basically creating one large encrypted file. With truecrypt you can then mount that file as partition. Your data is then encrypted/decrypted on the fly. You use is just like you would any other partition.

I created a 10gigabyte truecrypt partion on my linux laptop. That is about big enough to hold my personal data, bank statements, emails, pictures of the kids etc. When the truecrypt volume is not mounted, it looks like one big encrypted file. Once I start up truecrypt and put in my password, I then mount it and it works just like any other file partition. I have been using truecrypt for about six months with no problems. I can install software on the truecrypt partition and haven't noticed any speed difference.

So now the data is safe and encrypted on my laptop. Next problem is if I loose my laptop, I need a backup. My last check at Microcenter a DVD-R pack cost $35 for a 20 pack. That is about $.20 per gigabyte. Plus you have to store them in a place where the kids won't step on them.

Instead of dvds I decided to go with jungledisk and Amazon's S3. Amazon's Simple Storage Service (S3) is scalable on demand disk storage. The downside is Amazon only offers some low level APIs to access the storage. This is where jungledisk comes in. Jungledisk can connect to your Amazon S3 account and create a webdav frontend to your account. Once jungledisk was up an running I fired up davfs to mount my S3 account as a webdav drive. Jungledisk has on the fly encryption, so my files are stored encrypted on S3, so those nosey Amazon admins can't sneak a peak at my pictures.

So I have davfs -> jungledisk -> Amazon S3. It sounds more complicated than it is. In the end you have mounted S3 as a drive and basically have unlimited drive space.

Currently Amazon charges $.15 gig/month for storage and $.20 per gig bandwith. That makes is pretty competitive with dvd backups. My 10 gigabytes of data will cost $2.00 to upload to S3, then $1.50 month for storage. I use rsync to update the data every so often. Because only a small amount of my data changes each month, my bandwith charge is minimal. I end up paying about $2.00 per month for my backups.

One thing that I was a little worried about is that jungledisk is not open source. I didn't want to store my data encrypted then not be able to decrypt it. The developers of jungledisk have released a GPLd download/decrypt only version of jungledisk. So even if jungledisk went away tomorrow, I would still be able to at least get my data back.

A few things to note. On linux you have to set a few environment variables to make a truecrypte/ext2 partition over 1 gig. When using rsync with jungledisk, you need to use a few special flags.