Today I moved fast and broke things -- this site to be specific -- but it's all OK because I know what went wrong and (I think) I know how to fix it!

Here's the short story: I currently host this site on Amazon EC2, but since it's just a static site, it's hard to justify the cost of EC2, especially since there are much cheaper options that still allow me to manage my TLS certificates (which is a non-negotiable requirement for me). While I thoroughly enjoyed running my own server and tweaking my Nginx configuration, to be juuuust right, there are cheaper options that are better suited for my needs.

My new plan is to host this site's files on Amazon S3, and set up a CloudFront distribution that points to my S3 bucket, and upload a Let's Encrypt certificate to AWS' certificate management system. It seems like not much could go wrong right? Wrong. So many things went wrong, but I had a lot of fun troubleshooting and I learned a lot about A records and CNAMEs and DNS caching a whole host of other things yesterday.

I also spent some time rewriting my cryptopals solutions in Python and I'm much happier with the quality of my code compared to the solutions I wrote over a year ago. I also learned about Moore's voting algorithm and implemented it in Python and read more of The TCP/IP guide. Despite bringing down my site, it was a very productive day and I'm content with what I learned.

 Previous:  Recurse Center Day 6: Raft's Safety Argument and Cryptopals Round 2