Docker Container Lifecycles – Problem or Opportunity?
summary
Shows that Docker builds are usually nonreproducible so promotion of built images beats rebuilding at each stage, with a concrete claim that audience familiarity drops from ~90% heard Docker to ~5% running it in production. Presents patterns that work: use image digests not tags, host-per-stage registries via Artifactory plus proxying, promote artifacts with REST API, and sandwich testing to avoid framework/app deadlocks. Demonstrates practical fixes and tradeoffs.
AI-written summary · about
chapters (18)
- 0:00Start
- 0:12Patterns and Antipatterns in Docker Image Lifecycle
- 1:10Poll Time!
- 3:24JFrog Artifactory + Docker
- 4:56DO WE HAVE AN EXISTING PATTERN?
- 5:50The Promotion Pyramid
- 6:51Pipeline: Quality Gates and Visibility
- 9:00FAST AND CHEAP BUILDS
- 11:05Let's fix it!
- 16:54Traditional Server Pattern
- 17:08Immutable Server Pattern
- 19:34The Anatomy of Docker Tag
- 21:44Virtual hosts/ports to the rescue
- 24:20Anatomy of a container
- 25:47Minimal Framework build Dockerfile
- 26:45Application build Dockerfile
- 29:26Framework Pipeline
- 32:07Conclusions: Release Fast or Die!
related talks
41mDocker Container Lifecycles, problem or opportunity?
16mBest Practices for Container Image Promotion Pipelines
15mKeynote: Best Practices In Implementing Container Image Promotion Pipelines
40mBest Practices In Implementing Container Image Promotion Pipelines- 40mBest Practices In Implementing Container Image Promotion Pipelines
57mKiller Docker Workflows for Development
49mContinuous Delivery at Docker age
51mProduction level Docker