Choosing a non-relational database; why we migrated from MySQL to MongoDB

Server Density Blog

Update 23 Oct 2010: A discussion about the value of shortened field names has generated a lot of traffic to this post over the last 24 hours. A response has been posted here.

Update 28 Feb 2010: A followup was published here.

Until recently, our server monitoring application, Server Density, was running using MySQL for the backend. Although we primarily provide it as a hosted service, it has been written to work as a standalone application for customers that wish to install on their own servers. This means each customer had their own MySQL database.

We collect a lot of data – the monitoring agent reports back every 60 seconds and includes various statistics, of which the server snapshot has the most data (because it is collecting details on every running process). Over time, this results in millions of rows in the database, even for…

View original post 1,469 more words