Now that the Rails blog is back up, let me explain what happened. We upgraded our kernel and then upgraded Fedora from FC4 to FC5 on one of our servers, our other being Ubuntu. On both servers, we upgraded apache to apache 2.2 with mod_proxy_balance so that we could accomplish Load Balancing for our mongrel cluster setup for our rails apps (no more "Application errors") we hope, and well.. Upgrading Fedora broke quite a few system components and the mail server, so that had to be repaired also.. All of this in light of working full-time schedules during the day!
Oh, and I ought to mention that we upgraded to Typo 4.0.3 which didn't want to play nicely with Rails 1.1.6. To solve the problem, we deleted the sparklines textfilter from components/ and it seemed to fix the errors we were having. I wish I could say there was a perfect tutorial for setting up apache 2.2 with a mongrel cluster and load balancing from start to finish. A lot of people probably saw Coda Hale's tutorial and figured it out from there. We ran into some problems with it but for the most part it guided us in the direction we needed to go with things to figure the rest out for ourselves. We'll post an Apache 2.2/mongrel cluster/rails app tutorial so you can see how we did it as well. One thing to point out about this setup, is that it allows for apache to serve files and for mongrel to handle the rails application without being overwhelmed with non-rails related items such as images and static content. Let mongrel do what mongrel is best at and apache (which is written in C and very fast) serve the main site files. Having done some very basic benchmarking, we believe Ruby on Rails Blog is now 10-15% faster than it was before.