Amazon's S3

I’ve been experimenting with Amazon’s S3 web service, something that’s a sort of cross between a “remote hard drive” and a “remote database.” I’m testing it as a possible repository for a client’s nightly backup of about 40GB of data and scripts, and so far Christopher Shepherd’s PHP scripts (which use Geoff Gaudreault’s PHP S3 class to talk to S3) are the leading contender for a useful S3 toolkit.

I also did a lot of experimenting with s3sync, an rsync-like S3 backup solution in Ruby, but ran into some crashing problems seemingly related to some HTTP timeout problems in Ruby itself, so I left it aside for now.

I’ve enjoyed my thrash through S3 so much that I created an episode of The 3LA Podcast about it. If all of the above is Greek to your eyes, try giving it a listen — I attempted to sum it all up in rather less buzzwordy language in the podcast.