Modern Continuous Delivery
summary
Claims that continuous delivery requires a single reproducible artifact built once, deployable by clicking a button, and teams organized as product owners owning small independently deployable services. Shows tools and patterns: bash deployer Conan, GoCD spinning up 40 boxes to run 12,000 tests in eight minutes, Docker images pushed to registries, Kubernetes with Helm for declarative deploys, Vault and Twistlock for secrets and runtime security, and recommends canary deployment and testing observability data in staging and production.
AI-written summary · about
chapters (14)
- 0:00Introduction to modern CD
- 0:44Act one Solaris cluster lessons
- 4:19Act two traditional CD pipeline
- 7:48Act three microservices and infrastructure
- 8:29Modern continuous delivery today
- 13:27DevOps and team structure
- 15:35Trunk based development
- 16:39Testing in distributed systems
- 18:26Deployment strategies
- 21:59Dynamic cloud environments
- 22:42Observability over monitoring
- 24:07Security in DevOps
- 26:55Future of team ownership
- 29:03Conclusion and resources
related talks
19mModern Continuous Delivery
41mAll Things Continuous
24mWhy, What, and How of Continuous Delivery
42mCONTINUOUS DELIVERY FOR RUBYISTS
38mContinuous Delivery, or the magic button for on‑demand releases
26mWhere Are We Going? The Future of Continuous Delivery
28mMinimum Viable Continuous Delivery
40mMicroservice Continuous Delivery