An on-premises ecommerce app consists of Python microservices running in Docker with configuration via environment variables. You need a serverless GCP deployment. What should you do?

The correct approach is to use your CI/CD pipeline to build Docker images and deploy them to Cloud Run. Cloud Run is ideal for containerized microservices, offering a serverless platform that scales automatically and supports custom Docker images. Since the existing application uses Docker and environment variables for configuration, Cloud Run provides a natural migration path. You'll need to update configurations and endpoints to align with the GCP environment. Deploying to Cloud Functions is incorrect because Cloud Functions are designed for event-driven, single-purpose functions, not general-purpose microservices running in Docker containers. Deploying each service as a separate Cloud Function from the existing codebase would require significant refactoring, as Cloud Functions typically expect source code, not Docker images. Reusing on-premises configurations directly with Cloud Run is unlikely to work without updates, as GCP services have specific ways of handling secrets, environment variables, and service discovery.