Google Cloud

Google Cloud Platform (GCP)

Core Capabilities

Provides expert guidance for Google Cloud Platform across all major services:

1. Compute Services - Compute Engine (VMs), Cloud Run, Cloud Functions, App Engine
2. Container & Kubernetes - Google Kubernetes Engine (GKE), Artifact Registry, Cloud Build
3. Storage Services - Cloud Storage (buckets), Persistent Disk, Filestore
4. Database Services - Cloud SQL, Cloud Spanner, Firestore, Bigtable, Memorystore
5. Data & Analytics - BigQuery, Dataflow, Dataproc, Pub/Sub, Composer
6. Networking - VPC, Cloud Load Balancing, Cloud CDN, Cloud Armor, Cloud Interconnect
7. Security & IAM - Identity and Access Management, Secret Manager, Cloud KMS, Security Command Center
8. 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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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](references/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

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: DauQuangThanh
• Source: DauQuangThanh/hanoi-rainbow
• License: MIT

Install and usage instructions live in the source repository linked above.