An operations team is preparing to deploy an updated revision of a web application on Google Cloud. To minimize risk, they want to route a small portion of live user traffic to the new revision first to observe error rates and latency before proceeding with a complete rollout. Which deployment strategy best fulfills this requirement?
- Canary deployment with incremental traffic splittingAnswer
- BBlue-green deployment with an immediate 100% cutover without verifying schema backward compatibility
- CRe-architecting the simple stateless service to a dedicated Google Kubernetes Engine (GKE) cluster for traffic management
- DSubmitting a regional compute quota increase request before initiating any traffic migration
Answer
Canary deployment with incremental traffic splitting is the correct choice because it exposes the new software version to a small subset of production traffic to test operational metrics before full rollout.
A canary deployment routes a small percentage of live production traffic (such as 5% or 10%) to a newly released application revision while keeping the majority of traffic on the stable revision. This allows monitoring metrics like latency and HTTP error rates under real conditions before completing the rollout.
Step-by-Step Solution
Key Concept
Canary Deployment Strategy