This Site is Now Entirely on S3
Okay, so I got tired of the MongoHQ server connection issues that caused exceptions on my EC2 instances, and since this particular instance was running only this blog, I decided to make a change. I have been curious about the potential for hosting a site on Amazon Web Services’ Simple Storage Service (S3) since their announcement earlier this year of support for serving an index.html, if present when just a directory is requested, e.g. www.joshdmiller.com returns www.joshdmiller.com/index.html.
As this blog is relatively simple, I thought it would be a prime candidate to test this. Using Jekyll, a static site generator written in Ruby, I rewrote the code for this site, which involved only a few major steps: (1) seting up the basic infrastructure of the site; (2) translating the SCAML templates I used in Scalatra back into HTML; (3) translating the dynamically-generated content into its static equivalents. The posts were already in Markdown format, so nothing was really needed here, save some YAML Front Matter, per Jekyll’s docs. All said, this took only a few hours and I don’t have to pay for a server anymore. My verdict: Jekyll’s pretty cool.
I’ll throw the code up on a public repository within the coming days in case someone wants inspiration.