This is some background to my problem…
I have been using 3rd Party as a platform to host my rails application, and its great. So easy to get started and leverage the full power of AWS!
However for some of the tasks I wanted to do there were conflicts with gems / server / ruby versions that I could control locally through development, but never worked on my 3rd Party as there is a certain amount control over the environments you can’t change. Which means its a stable platform but you can’t really go outside the box. I am using Trinidad / Jruby which has great performance advantages!
This is what I am doing
So I decided to launch my own EC2 instance. This is going really well. However I am interested in knowing what everyone would do in terms of deployment. At the moment I am downloading the git into the EC2 server and running rails from command. Ideally I would like rails to run in the background, and launch at start up.
Question
How would you configure the EC2 instance to be a full on production server and not just a server run from command through SSH. Is there anyway to set up your own deploy tool for example?
Thanks
Charlie
After a lot of research and trial and error the solution I went for was Torquebox
It is a Jruby (rails) server based on JBOSS and it can do Messaging, Queues, background jobs all in one. You can also deploy to it and its easy to get it loaded as a upstart process.
Its no engineyard or heroku, but its a lot more stable than any other solution I tried. And its quick!