An on-prem analytics binary runs nightly for ~45 minutes, processing 1–16 GB files in memory. Migrate to GCP with minimal effort and cost. What should you do?

The best solution for minimal effort and cost is to lift and shift to a Compute Engine VM. The existing binary can run directly on a VM, requiring minimal code changes. An instance schedule ensures the VM runs only when needed, minimizing cost. Containerizing for Cloud Run or GKE introduces overhead for containerization and orchestration, which is unnecessary for a simple, scheduled job. Cloud Functions are designed for stateless, event-driven functions, and the 1-16 GB memory requirement is likely too large and complex for a typical Cloud Function.