Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud Architect practice bank has 2,335 more like it. Try 10 free: instant score, no account.
An enterprise runs a critical transaction processing API on Google Cloud. The development team is preparing to release a major version update that includes extensive refactoring of core business logic.
The organization has established the following technical and risk mitigation requirements for the release:
Which deployment strategy should the Cloud Architect recommend?
A canary deployment (also referred to as an AB-style or phased rollout) is a release strategy where a new version of an application is deployed alongside the existing stable version, and a small percentage of live production traffic is routed to the new revision while the majority remains on the baseline.
Compared to rolling or all-at-once strategies, a canary deployment offers the best balance between validating critical code updates under real production loads and minimizing organizational risk.
A rolling update is an in-place deployment strategy that progressively updates instance replicas (such as Kubernetes Pods or Compute Engine Managed Instance Group instances) in batches until all instances are running the new release.
While a rolling update prevents downtime by incrementally replacing old Pods with new ones, it does not support precise, decoupled percentage-based traffic routing. All Pods in a shared Service pool receive load-balanced traffic proportionally, making it difficult to isolate a tiny sample of live users. Furthermore, rolling back requires executing a reverse rolling update across the cluster, which takes time and cannot instantly stop defect exposure.
The requirement was to perform routine, backwards-compatible service updates with minimal infrastructure overhead and without needing separate traffic-routing controls or isolated percentage canary phases.
A blue-green deployment maintains two separate, identical physical or virtual production environments: one running the active version (Blue) and one running the target new version (Green). Once testing in the idle environment is complete, traffic routing is switched over completely at the load balancer or router.
Blue-green deployments switch 100% of production traffic from the old environment to the new environment all at once. This fails the requirement to limit the initial blast radius by routing only a small, fractional percentage of production traffic to the new revision for telemetry verification before full release.
The workload required instant rollback capabilities and could not tolerate running two different API versions side-by-side processing live user requests simultaneously.
A recreate deployment is a strategy where all running instances of the existing application version are completely terminated before new instances running the updated version are launched.
A recreate deployment introduces explicit service downtime between the shutdown of the old instances and the startup of the new ones. It also lacks fractional traffic routing, telemetry validation against live users, and seamless rollback mechanisms, directly violating the zero-downtime requirement.
The application used a legacy datastore schema that strictly prohibited concurrent access by multiple application versions and scheduled maintenance windows with acceptable downtime were permitted.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated