The road to continuous deployment: a case study
summary
A legacy job-board monolith is cut down using the strangler pattern and an API-first service architecture deployed in Docker containers; CI/CD is implemented as Jenkins pipeline-as-code with Ansible and rolling updates, health checks and feature toggles, and trunk‑based development with pair programming. Result: build-to-production per service is under 10 minutes, the team ships 50+ deploys per day, page loads fall from ~6s to ~0.5s, outages and weekend releases drop sharply and team velocity and confidence increase.
AI-written summary · about
chapters (26)
- 0:00Intro
- 0:38Case study
- 4:15Goals
- 7:28The approach
- 9:46Architecture
- 10:56Rules
- 11:22Continuous
- 12:01Continuous Deployment
- 16:18Commits to production
- 17:31No branch
- 18:06Early integration
- 19:50Feature toggles
- 22:00Code reviews
- 23:44The Boy Scout rule
- 24:31Quality
- 24:53Code Coverage
- 26:06DevOps
- 26:50Team responsibility
- 27:34Build pipelines
- 28:30Continuous testing
- 30:34The testing pyramid
- 32:26Pipelines
- 33:06Rolling Update
- 34:55Siren of Shame
- 36:42Results
- 41:46Literature
related talks
53mThe Road to Continuous Deployment: a Case Study
01h 02mThe Road to continuous deployment
34mEnabling a Continuous Deployment Mindset
43mContinuously deploying a distributed monolith
01h 01mA team's transition to Continuous Delivery
23mDevOps and Legacy Code
33mThe Road to Enable DevOps Within the Payment Industry
12mContinuous Deployment in a heavily regulated environment