This is a snippet from a 2004 article that I never publicly published, but is serves as a good case study relevant in 2011 (and beyond).
In 2004, weather.com site served more than 50 million pages on stormy days, and it ran almost entirely on open-source software and commodity hardware. The Atlanta-based Web site’s adoption of a new architecture and open source products “has slashed IT costs by one-third and increased Web site processing capacity by 30%” (King 2004). However cost slashing was not their primary goal of switching to an open source product. The quality of open source products was its main “selling” point. Weather.com claimed that their transition from IBM’s server software product to open source Apache Tomcat to run their website served correct operations, ease of use and better quality attributes overall. Of course, there are different organizational dynamics that lead to a decision to drop COTS (and support) to an open source solution.
Performance and scalability issues were cited as the main reasons for switching to Apache’s web server. The team switched from IBM’s commercial offering to Apache’s open source implementation primarily for its quality. Apache’s open source web servers host 68% of web servers in the world according to an August 2004 analysis of Netcraft (Gustafson, Koff).
IBM has since started to use a modified version Apache Web Server in it’s commercial offerings. This is a trend that is likely to repeat itself across multiple technology domains depending on various factors: it remains to be seen if the penetration projections hold true over time.