Google Cloud Platform (GCP)
Core Capabilities
Provides expert guidance for Google Cloud Platform across all major services:
- Compute Services - Compute Engine (VMs), Cloud Run, Cloud Functions, App Engine
- Container & Kubernetes - Google Kubernetes Engine (GKE), Artifact Registry, Cloud Build
- Storage Services - Cloud Storage (buckets), Persistent Disk, Filestore
- Database Services - Cloud SQL, Cloud Spanner, Firestore, Bigtable, Memorystore
- Data & Analytics - BigQuery, Dataflow, Dataproc, Pub/Sub, Composer
- Networking - VPC, Cloud Load Balancing, Cloud CDN, Cloud Armor, Cloud Interconnect
- Security & IAM - Identity and Access Management, Secret Manager, Cloud KMS, Security Command Center
- Infrastructure as Code - Terraform Google provider, Deployment Manager, Config Connector
Key Principles
General Best Practices
- Follow least privilege - Use IAM roles with minimal required permissions
- Enable monitoring - Configure Cloud Monitoring and Logging for all services
- Use managed services - Prefer GKE Autopilot, Cloud SQL, Cloud Run over self-managed
- Implement IaC - Use Terraform or Deployment Manager for reproducible infrastructure
- Tag resources - Apply labels for cost allocation and organization
- Design for HA - Use regional resources and multi-zone deployments
- Secure by default - Enable encryption, use private IPs, configure VPC Service Controls
- Optimize costs - Use committed use discounts, autoscaling, and appropriate resource sizing
Architecture Patterns
- Multi-tier applications: VPC + Cloud Load Balancer + GKE/Cloud Run + Cloud SQL
- Data pipelines: Pub/Sub → Dataflow → BigQuery with Cloud Storage staging
- Serverless APIs: Cloud Run + Cloud SQL + Secret Manager + Cloud Armor
- Hybrid connectivity: VPN or Cloud Interconnect + Shared VPC + Private Google Access
When to Use What
- Compute Engine: Full VM control, Windows workloads, lift-and-shift migrations
- GKE: Containerized applications, microservices, Kubernetes workloads
- Cloud Run: Stateless HTTP services, event-driven processing, auto-scaling needs
- Cloud Functions: Event handlers, webhooks, simple integrations
- Cloud SQL: Relational databases with minimal management
- Cloud Spanner: Global distributed SQL, strong consistency across regions
- BigQuery: Data warehouse, analytics, large-scale SQL queries
- Firestore: Document database, real-time sync, mobile/web apps
Detailed References
Load reference files based on specific needs:
- Compute Services: See compute-services.md for:
- Compute Engine machine types and selection guide - Managed instance groups and autoscaling - Custom images and startup scripts - Preemptible VMs and spot instances
- Container Orchestration: See container-orchestration.md for:
- GKE cluster setup and configuration - Autopilot vs Standard mode comparison - Node pool management and scaling - Workload identity and service accounts - GKE Ingress and Gateway API
- Storage Solutions: See storage-solutions.md for:
- Cloud Storage bucket configuration - Storage class selection and lifecycle policies - Persistent disk types and performance - Filestore for shared file systems
- Database Services: See database-services.md for:
- Cloud SQL instance configuration - Cloud Spanner for global databases - Firestore data modeling - Bigtable for large-scale NoSQL - Memorystore for Redis/Memcached
- Data & Analytics: See data-analytics.md for:
- BigQuery table design and optimization - Dataflow streaming and batch pipelines - Pub/Sub messaging patterns - Cloud Composer (Airflow) workflows - Data governance and security
- Networking Architecture: See networking-architecture.md for:
- VPC design patterns and subnet planning - Cloud Load Balancing configuration - Cloud CDN and Cloud Armor setup - VPN and Cloud Interconnect - Shared VPC and peering
- Serverless Computing: See serverless-computing.md for:
- Cloud Functions deployment and triggers - Cloud Run service configuration - App Engine standard and flexible - Event-driven architectures - Cold start optimization
- Security & IAM: See security-iam.md for:
- IAM roles and service accounts - Organization policies and constraints - VPC Service Controls - Secret Manager integration - Cloud KMS encryption - Security Command Center alerts
- Infrastructure as Code: See infrastructure-as-code.md for:
- Terraform Google provider patterns - Deployment Manager templates - Config Connector for GKE - CI/CD with Cloud Build - State management best practices
- Migration to GCP: See migration-to-gcp.md for:
- Migration planning and assessment - Migrate for Compute Engine (Velostrata) - Database migration service - Storage transfer service - Cutover strategies and validation
- Monitoring & Logging: See monitoring-logging.md for:
- Cloud Monitoring setup and metrics - Cloud Logging configuration - Log-based alerts and metrics - Cloud Trace for distributed tracing - Cloud Profiler for performance - Dashboards and SLO monitoring
- CI/CD Pipeline: See cicd-pipeline.md for:
- Cloud Build configuration - Artifact Registry for containers - Deployment to GKE, Cloud Run, App Engine - Binary Authorization for security - Integration with GitHub, GitLab
- Cost Management: See cost-management.md for:
- Billing reports and cost allocation - Budget alerts and quotas - Committed use discounts planning - Resource optimization strategies - Cost anomaly detection
- Multi-Region Architecture: See multi-region-architecture.md for:
- Global load balancing patterns - Multi-region database replication - Cross-region data transfer - Disaster recovery strategies - Regional failover setup
- Hybrid & Multi-Cloud: See hybrid-multi-cloud.md for:
- Anthos for hybrid Kubernetes - Cloud Interconnect and VPN - Multi-cloud networking patterns - Workload migration strategies - Identity federation
- GCP CLI & Tools: See gcp-cli-tools.md for:
- gcloud CLI installation and configuration - Common gcloud commands - Cloud Shell usage - gsutil for Cloud Storage - bq for BigQuery operations - kubectl for GKE management