Inspired by this account from Jeremy Zawodny, I've moved to using Amazon S3 for all my backup needs.

Why did I do this?

I have around 7 GB of 'stuff'. Everyone has this cruft that accumulates over the years - code, mail, photos, etc. Most of it is of little value but you don't want to throw it away. I usually access this from multiple machines - my work laptop, my office dev boxes and my home computer. It's a pain in the rear to keep them all in sync.

I use SyncToy to keep them all in sync but that was a patchy solution - I needed a long term backup solution, especially given that I don't live in a 'permanent address' right now.

How I used S3

After reading Jeremy's account, I went and signed up for a AWS account and used my credit card to get into the S3 billing system. After that, it was a simple matter of using s3sync and a batch file to upload all the data I cared about to Amazon's machines.

The good

• Great billing and pricing scheme.


• Great API - which has lead to an explosion of tools


• Accessible through multiple mechanisms - Http, Bittorrent,etc


• The ability to put ACLs at very granular level. This combined with the last point means that I can expose one photo alone to the whole world and give people a URL to get to it

The not-so-good

• Inherently non-hierarchial structure means that every tool writer has his or her own idea of how a filesystem should be mapped onto the objects and key/value system. This leads to each tool being incompatible with every other tool out there. If you want to get your content easily, you might be limited to using the tool you used to upload your content in the first place.
  



• No resume for broken upload operations. Since upload bandwidth is limited on most broadband connections, this is a major pain. I have a 1 GB PST file which I still haven't managed to upload even after multiple attempts. Yes, I know I can split it up but that's besides the point