Cutting edge Continuous Delivery: Automated Canary Analysis through Spring based Spinnaker
summary
Spinnaker provides a Spring‑based CD platform that runs automated canary analysis by spinning up baseline+canary instances, pulling artifacts from S3/Travis, and evaluating Datadog metrics (CPU, JVM memory, rendering time avg/max, logback events) on configurable intervals and warmups. Netflix uses a 95% threshold and multi‑hour canaries; the demo runs a 6‑minute canary, observes a ~190% spike in rendering time for the canary, scores it as failure, and automatically tears down the canary and restores the baseline.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:08The challenges of CD
- 4:35Introduction to Spinnaker
- 8:03Architecture and ecosystem
- 11:53Deployment best practices
- 17:09Advanced deployment strategies
- 24:24Canary analysis and safety
- 31:08Live demo: Rick and Morty
- 47:52Conclusion and Q&A
related talks
35mCutting-edge Continuous Delivery: Automated Canary Analysis on CF using Spinnaker
35mReduce Risks With Continuous...
42mMetrics Driven Blue-green Deployments using Spinnaker’s Cloud...
52mContinuous Deployment to the Cloud using Spinnaker
15mYahoo Case Study: Automated Canary Analysis using Spinnaker Kayenta
43mCanary Deploys with Istio: Lessons Learned
12mImplementing Advanced Deployment Using Spinnaker and Canary
34mEnabling Invisible Infrastructure Upgrades with Automated Canary Analysis