Yahoo Case Study: Automated Canary Analysis using Spinnaker Kayenta
summary
Yahoo/Verizon Media runs Spinnaker Kayenta in Kubernetes integrated with Screwdriver to perform automated canary analysis, extending Kayenta with Splunk and an OpenTSDB internal metrics plugin, a Redis cache, and S3-backed configs; traffic uses a 90/5/5 production/baseline/canary split and analyses run ~60 minutes for statistical confidence. A Terraform-style HCL-inspired DSL generates compact canary configs, a Screwdriver template automates pipeline steps, and Nike’s Referee provides a UI; solution is deployed and used for onboarding with plans for static-threshold judges and early-fail detection.
AI-written summary · about
chapters (3)
related talks
51mCutting edge Continuous Delivery: Automated Canary Analysis through Spring based Spinnaker
43mCanary Deploys with Istio: Lessons Learned
36mHow To Use Referee & Kayenta to Canary...
35mReduce Risks With Continuous...
34mEnabling Invisible Infrastructure Upgrades with Automated Canary Analysis
35mCutting-edge Continuous Delivery: Automated Canary Analysis on CF using Spinnaker
52mContinuous Deployment to the Cloud using Spinnaker
24mUsing Jenkins and Spinnaker to Supercharge Your Deployment Pipelines