What are the architecture details of Stack Overflow?

Question

Software Stack: https://blog.stackoverflow.com/2008/09/what-was-stack-overflow-built-with/

Database Details: https://blog.stackoverflow.com/2009/06/stack-overflow-creative-commons-data-dump/

What are the details of the Stack Overflow development architecture. Things such as schemas and queries related to how they handle tagging, caching of various parts of the page, how the database is configured to handle greater number of reads to writes (or vice versa if it's write heavy).

If these can't be answered directly regarding the exact implementation, then answers or discussion on how you would implement the different aspects of the site would be great, or Best Practices regarding how to implement a site similar to this. Also, maybe a collection of the links to the posts that describe an aspect of the site design.

• Caching
• SQL Server Configuration
• Tagging

I have updated the question to make it more clear what I was looking for. TheTXI answered with a link to the different technologies and info on the what the software stack consists of.

If you still think this is a duplicate or irrelevant question, please edit it or remove it.

• Pops: possible duplicate of Which tools and technologies build the Stack Exchange Network?

Answer 1

This has actually been asked before and has been covered in the StackOverflow blog (and I am sure somewhere in the FAQ).

Here is the link to the blog post:

https://blog.stackoverflow.com/2008/09/what-was-stack-overflow-built-with/

• Jeff Atwood: I think there's a nice post on SO someone put together on this too.. should be migrated here once the Google cross-site account associations are up and running.
• Shog9: The SO FAQ entry: stackoverflow.com/questions/749358/… Note that it's been deleted for almost a month - good opportunity for someone to repost it here...
• Steve T: I was sort of hoping for a discussion on how they implement tags, caching, and possibly things like how the site is designed. I understand this probably won't come from the team themselves as they aren't open sourcing the project yet.
• Pops: I have voted to close as a dupe of the post I think Jeff is talking about.

Answer 2

There is a database export available, this will give the database structure.

J&J talk about some implementation details on the podcast from time to time (e.g. the database refactoring to move current revisions of questions and answers to avoid joins.

• Steve T: This is useful thanks, maybe I'll have to compile all the info from the podcasts, as I know I've heard them talk about it, but can't remember everything.