Moving my online-presence to another server, I now have a lot less space to store my backups on than I had before. Thus, a simple rsync-via-ssh is not really an option any more. This forced me to look for another solution. Luckily, there is a great abundance of online backup services, most of which use Amazon S3 as the backend storage. I seemed to have two options to use S3 as the backup provider:

1. Use some commercial and complete software for making, transferring and organizing the backups. A couple of solutions seemed mature enough to real-world usage: Dropbox and Jungle Disk. Several user reports recommend both solutions, which should be quite stable and easy to use. Also, they are not that expensive compared to the raw cost of S3. Both commersial solutions have disadvantages, though – Dropbox adds a lot of cost if you want to store more than 2GB but a lot less than 50GB (compared to S3), and Jungle Disk apparantly stores the backed up files in some odd way, making it hard to recover files without Jungle Disk. Also, they share the disadvantage of being commercial (it is non-free and the source code is not available). An advantage of both solutions is that they are multiplatform.
2. Use some combination of open source software to back up my files directly to Amazon S3. The open source software available for communicating more or less directly with S3 seems quite mature at this time. Browsing the blogosphere gave me a lot of documentation on how to do this using varying tools ranging from duplicity through two (different) projects called s3fs and s3fs (??) to encFS. Using these tools, it is possible to customize the backup of files in exactly the way the user wants, and they do not present the user with any extra cost besides the raw cost for the S3 storage.

I chose the second option, using a bit more time in setting up the backup but in return getting a completely customized and free backup toolchain. In the next couple of posts I will document how I did it, but to get an idea (and to get some inspiration), here is the resources I used when setting up the backup:

http://blog.eberly.org/2008/10/27/how-i-automated-my-backups-to-amazon-s3-using-rsync/
http://www.cenolan.com/2008/12/how-to-incremental-daily-backups-amazon-s3-duplicity/
http://www.brainonfire.net/blog/remote-encrypted-backup-duplicity-amazon-s3/
http://sharph.net/2008/12/encrypted-offsite-backup-with-encfs-amazon-s3-and-s3cmd/
http://markusthielmann.com/blog/use_amazon_s3_with_ubuntu_hardy
http://www.sharms.org/blog/?p=137