The Future of Continuous Delivery
summary
Claim: a monolithic CI pipeline for a UK government project drops from 1.5 hours to 12 minutes by using heavy parallelism, one beefy build node to reuse images immediately, and test-suite segmentation. Tools actually used: Jenkins (classic), Tekton, Kubernetes/k3s, GitOps with Flux and Argo, Argo events for event-driven workflows, and monitoring with Prometheus/Dynatrace integrated into a Captain control plane. Future direction is small, event-emitting workflows that coordinate via these tools and can trigger automated rollbacks.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:15Talk structure
- 5:27Continuous delivery past
- 12:46Tekton cloud innovation
- 14:31Modern CD tools
- 15:55GitOps complexity
- 17:40Jenkins X solution
- 23:57Flux CD Kubernetes
- 27:07Argo CD workflows
- 29:35Captain control plane
- 39:13Continuous delivery future
- 41:18CD standardization
- 42:00Event-driven pipelines
- 45:10Q&A
related talks
32mEvent-based Continuous Delivery: The End of the Monolithic Pipeline
01h 02mContinuous Integration: Tips & Tricks
01h 01mA team's transition to Continuous Delivery
01h 26mOptimising Continuous Delivery
44mSimplifying and Scaling Software Delivery
41mAll Things Continuous
06mBuilding Continuous Delivery Pipelines
34mBeyond Continuous Delivery