Quantifying Tech Debt

2015 DevOps Days Boston - Lightning talk

Transcript

1. Quantifying Technical Debt Casting magic missiles at the darkness Anthony Spring DevOps Days Boston - 2015
2. Definitions Quantification - “express or measure the quantity of.” Technical debt - “work that needs to be done before a particular job can be considered complete or proper.”* https://en.wikipedia.org/wiki/Technical_debt
3. Googling: Types of Debt Architectural changes Design considerations Hard coded values Lack of documentation Redundant code Redundant code Test coverage
4. Quantification: Goals Ease of use Minimal friction — Getting developer buy in is hard Ability for developers to breadcrumb notes related to the perceived technical debt. Reporting —automatically generated per project
5. Jenkins Task Scanner Plugin Added three new task tags: YELLOW, ORANGE, RED Each color provides a rough estimate as to the time/ complexity of the technical debt. Allows for developer to add their own comments and considerations regarding the debt and paying it off. Not qualitative. We were already using it!
6. Text Jenkins Task Scanner Plugin* https://wiki.jenkins-ci.org/display/JENKINS/Task+Scanner+Plugin * I am not the author
7. Text Job View
8. Text Details View per File Easily see the developer breadcrumbs to help plan paying down the debt.
9. Text Dashboard View Providing an easily visualization of trending across projects
10. How do we use it? Review new debt Is it necessary? Is it classified correctly? Review old debt — was is fully paid off? Team OKR to pay off a specific amount of debt per quarter.
11. –Random Member of Audience “You talk too fast. Can you clarify ________?”
12. Anthony Spring Minister of DevOps Culture / Infrastructure Tech Lead aspring@yieldbot.com @devopsanthony