Google Cloud (GCP) lock-in assessment¶

Lock-in risks¶

• Proprietary Databases:

  • Cloud Spanner (globally distributed, unique SQL dialect)

  • Firestore (Google-specific NoSQL API)

  • BigQuery (serverless analytics with GCP-only optimizations)

• Serverless & Compute:

  • Cloud Functions (Google-specific triggers)

  • Cloud Run (managed containers with GCP quirks)

• Storage & Networking:

  • Cloud Storage (similar to S3 but different API)

  • Google’s global load balancer (unique to GCP)

• Management & IaC:

  • Deployment Manager (GCP-only)

  • Org Policies (Google-specific governance)

How to check for lock-in¶

1. Inventory: Use Anthos Config Management for dependency mapping.

2. API Audit: Check for @google-cloud/* libraries.

3. Cost Check: Review GCP Pricing Calculator (egress: $0.08–$0.12/GB).

Mitigation strategies¶

• Use standard SQL (PostgreSQL/MySQL) instead of Spanner.

• Replace Cloud Functions with Knative on GKE.

• Use Terraform instead of Deployment Manager.