Sorry for the delay in updating this series about Chef and Jenkins. Last week I was busy celebrating with family and went through the Master’s hooding and commencement ceremonies. Unfortunately, I didn’t get a chance to revisit the blog until now.
This week I was getting acclimated to my new job at Cycle Computing. I am super stoked about this opportunity, and all of my Cycle colleagues seem like a friendly, and smart, bunch of people.
I wanted to get a few more thoughts from my Masters Capstone project into blog form. I want others to be able to learn from both the technical and theoretical side of my research. This post is distilled continuous integration theory applied to Chef cookbooks. Its not prescriptive, and it may not work for you. It was extremely effective in practice at Marshall, however.
Over a period of 12 weeks, there were 159 changes introduced into two cookbooks. There was a nominal failure rate of 19.2% to 20.31% in the continuous integration environment. The more realistic effective failure rate was actually 16.43% to 18.37% in the continuous integration environment. However, there was only a 0.99% to 1.92% rate of defect release for the in-scope software units over the course of the project.