Canary Deployment

Release to a small percentage of traffic, monitor, then gradually increase.

deploy

Detailed Description

Canaries reduce release risk by exposing new code to a small group first.

Automated canaries should watch error rate, latency, and business metrics.

Roll out in traffic steps (for example 1% -> 5% -> 25% -> 50% -> 100%) with explicit gates.

Define rollback thresholds in advance: 5xx spike, p95/p99 latency regression, or checkout/payment drop.

Canary is a release-safety strategy, not a product experiment strategy.

Visual Diagram

100 users total:
  5%  → v2 (canary 🐦) — monitor closely
  95% → v1 (stable)

  metrics OK? → increase to 25% → 50% → 100%
  metrics bad? → rollback 5% → 0% instantly

Tradeoffs

Pros

Real-user validation with limited blast radius

Cons

Needs metrics and traffic splitting

Examples: Argo Rollouts, Flagger, Spinnaker, Istio weights