A financial services organization is automating its deployment pipeline using Google Cloud Deploy to manage releases across GKE clusters in staging and production. To ensure compliance and zero downtime during application upgrades, the release process must deploy new microservice revisions to a small percentage of user traffic first, validate service metrics, and support an automated instant rollback if error thresholds are exceeded. Which deployment and automation strategy should the Cloud Architect implement?
- Define a progressive canary deployment strategy in the Cloud Deploy delivery pipeline configuration, integrated with automated pipeline hooks to evaluate Cloud Monitoring metrics before promoting traffic.Answer
- BGrant the primitive Owner role to the Cloud Deploy execution service account and store the release state in an unversioned local directory within the CI worker container.
- CConfigure the pipeline to grant Service Account Admin rights to the Cloud Build agent so it can provision GKE clusters on-demand for every application release step.
- DStore all Terraform environment states locally on the build worker and rely on manual Cloud Console resource updates whenever a canary stage fails validation.
Answer
Define a progressive canary deployment strategy in the Cloud Deploy delivery pipeline configuration, integrated with automated pipeline hooks to evaluate Cloud Monitoring metrics before promoting traffic.
Google Cloud Deploy natively supports progressive deployment strategies such as canary releases for GKE and Cloud Run. By defining canary percentage phases in the delivery pipeline configuration and integrating verification hooks, organizations can automatically validate application metrics and perform automated rollbacks if issues occur.
Step-by-Step Solution
Key Concept
Automating Continuous Deployment Pipelines and Progressive Release Strategies using Cloud Deploy
Estimated Time:1m 30s