S3's Super Backups

Published on Monday, January 22, 2007

My buddy Ian  mentioned Amazon's S3 service, and the potential for using it for fun webapps.  While utilizing it for webapps will have to wait a few months, I was able to use it as a cheap backup for my home server (pictures, documents, etc,.) -- and my server that houses my email and websites.  The setup is pretty quick, and most of it can be detailed here.  The ruby package is here   I'll toss in my recommendation to use the jets3t Cockpit application for viewing the buckets, especially considering the Firefox extension didn't work as advertised.  My only two comments will be this:

1) Making sure SSL is working.  The site mentioned above just has you hunt down some random bash file, that isn't even hosted anymore.  On my Debian system I simply added this to my upload.sh:

export SSL_CERT_DIR=/etc/ssl/certs/

2) The second suggestion is another example of the s2sync layout.  Let's say you created the bucket "kelvinism" -- the following would move the documents inside a test folder from /home/kelvin named test to a folder named test in the kelvinism bucket.  Sweet.

 s3sync.rb -r --ssl --delete /home/kelvin/test kelvinism:/test