Argo Rollouts is a Kubernetes controller and set of CRDs that adds advanced deployment capabilities — blue-green, canary, canary analysis, experimentation, and progressive delivery — to Kubernetes clusters. It acts as a drop-in replacement for the native Deployment object when more sophisticated traffic control and metric-driven promotion or rollback is needed. The key mental model: Argo Rollouts separates deployment (updating pod specs) from traffic shifting (routing percentages to versions), enabling risk reduction at every release by limiting the blast radius of a bad change before it reaches all users.
What This Cheat Sheet Covers
This topic spans 17 focused tables and 118 indexed concepts. Below is a complete table-by-table outline of this topic, spanning foundational concepts through advanced details.
Table 1: Core CRDs and Architecture
Argo Rollouts introduces five custom resources that together form the progressive delivery engine. Understanding what each resource does and how they relate is the foundation for working with any strategy or analysis feature.
| Resource | Example | Description |
|---|---|---|
apiVersion: argoproj.io/v1alpha1kind: Rollout | • Drop-in replacement for a Kubernetes Deployment• adds blueGreen and canary strategies plus analysis integration | |
kind: AnalysisTemplatespec: metrics: - name: success-rate | • Defines how to measure analysis — which metrics to query, their interval, and pass/fail thresholds • reusable across Rollouts | |
kind: AnalysisRunstatus: phase: Successful | • Live instantiation of an AnalysisTemplate• runs like a Job and concludes as Successful, Failed, or Inconclusive. |