Writing a Ractor-based web server: part II

A few months ago I published a post about writing a simple web server in Ruby using Ractors. That took only 20 lines of code and it was able to leverage multiple CPUs with Ruby without having to go through the Global Interpreter Lock (GIL). That was a good preview to what the Ractor primitive is going to provide.Since then Ruby 3.0 was released and the Ractor implementation has got more mature. In this post, we’ll make our Ractor-based web server do more things.By the end of the post, you’ll learn the constraints of Ractors and get familiar with three PRs to MRI that I had to open to make it work.