Google Cloud Platform (GCP) - Practical Handbook
Google Cloud is easiest to learn from projects, IAM/service accounts and VPC, then choose the runtime layer that matches the workload: Compute Engine, Cloud Run or GKE. For a simple container, Cloud Run is usually a much easier starting point than Kubernetes.
Google now refers to its serverless function product as Cloud Run functions; the Cloud Functions name still appears in older APIs and migration material.
Related topics: Computer Networks for Developers, Docker, SQL and PostgreSQL, Web Analytics and Tagging and CI/CD and GitHub Actions.
1. What Google Cloud is
Google Cloud is a public cloud platform offering compute, storage, databases, analytics, networking, containers, serverless and AI services.
2. Resource hierarchy
organization
↓
folders
↓
projects
↓
resources
Projects are central billing, IAM and API-management boundaries.
3. Regions and zones
Regions are geographic locations.
Zones are separate infrastructure locations inside a region.
4. Core services - map
Compute Engine virtual machines
VPC networking
Cloud Storage object storage
Persistent Disk block storage
Filestore managed file storage
Cloud Run managed serverless containers
GKE Kubernetes
Cloud Run functions
Artifact Registry images/packages
Cloud SQL relational databases
Spanner distributed relational database
Firestore document database
BigQuery analytics warehouse
Pub/Sub messaging
IAM permissions
Service Accounts workload identities
Secret Manager secrets
Cloud Logging logs
Cloud Monitoring metrics
Vertex AI AI/ML platform
5. Projects and APIs
Many Google Cloud services require enabling a project API before use.
List enabled services:
gcloud services list --enabled
6. Compute Engine
Classic virtual machines.
Important concepts:
- machine type,
- image,
- persistent disk,
- VPC/subnet,
- firewall rules,
- service account,
- metadata.
7. VPC
Google Cloud VPC networks are global resources with regional subnets.
Core concepts:
- VPC,
- subnet,
- routes,
- Cloud NAT,
- peering,
- private service access.
8. Firewall
Firewall rules control traffic to VM workloads.
Use tags/service accounts and narrow source ranges where possible.
9. Cloud Storage
Object storage using buckets and objects.
Useful for static assets, backups, archives and data pipelines.
10. Persistent Disk and Filestore
Persistent Disk: block storage for VMs.
Filestore: managed shared file storage.
11. Cloud Run
One of the simplest GCP services for small containerized applications.
You provide a container image and Cloud Run handles scaling and HTTPS endpoint exposure.
12. GKE
Google Kubernetes Engine is managed Kubernetes.
Use it only when Kubernetes capabilities justify its operational complexity.
13. Cloud Run functions
Event-driven function execution.
Modern Google Cloud function products are increasingly integrated with Cloud Run infrastructure.
14. Artifact Registry
Stores container images and software packages.
15. Cloud SQL
Managed PostgreSQL, MySQL and SQL Server.
16. Spanner
Globally scalable distributed relational database.
Powerful, but usually unnecessary for small projects.
17. Firestore
Managed document database often used by web/mobile applications.
18. BigQuery
Serverless analytical data warehouse.
Designed for large-scale SQL analytics.
19. Pub/Sub
Managed messaging service for asynchronous/event-driven architectures.
20. IAM
IAM controls who can do what on which resource.
Prefer narrow roles over broad primitive roles.
21. Service Accounts
Workload identities for applications and machines.
Prefer attached service accounts and short-lived credentials over downloaded static keys.
22. Secret Manager and KMS
Secret Manager stores sensitive configuration.
Cloud KMS manages encryption keys.
23. Cloud Logging and Monitoring
Use these for:
- logs,
- metrics,
- dashboards,
- alerts,
- service troubleshooting.
24. gcloud CLI
Login:
gcloud auth login
Show configuration:
gcloud config list
Select project:
gcloud config set project PROJECT_ID
25. Terraform
Terraform is widely used for GCP infrastructure as code.
Google also provides native deployment/configuration tooling.
26. Load Balancing, CDN and DNS
Cloud Load Balancing supports global and regional architectures.
Cloud CDN provides content delivery.
Cloud DNS manages DNS zones.
27. Vertex AI
Managed AI/ML platform for model training, deployment and generative-AI services.
28. Example architectures
Small Go application
Cloud Run
→ Cloud SQL PostgreSQL
→ Secret Manager
Data pipeline
Pub/Sub
→ Dataflow / processing
→ BigQuery
Kubernetes
Load Balancer
→ GKE
→ managed databases/services
29. Costs
Watch:
- idle Compute Engine,
- Cloud SQL size,
- BigQuery scans,
- log volume,
- network egress,
- GKE clusters,
- retained disks/snapshots.
30. Project organization
Use separate projects for meaningful environment or ownership boundaries.
Examples:
app-dev
app-prod
shared-network
data-platform
31. Security baseline
- MFA for users,
- least privilege IAM,
- service accounts for workloads,
- avoid static service-account keys,
- Secret Manager,
- private networking where useful,
- audit logging,
- budgets.
32. Common traps
- using one project for everything,
- leaving broad Owner/Editor roles,
- exposing Cloud SQL publicly,
- unnecessary GKE,
- ignoring BigQuery/logging costs.
33. Small-project choices
Cloud Run is often the first service to evaluate for small containerized apps.
Compute Engine is still useful when you want a classic VPS model.
34. CLI cheat sheet
gcloud auth list
gcloud config list
gcloud projects list
gcloud compute instances list
gcloud compute networks list
gcloud storage buckets list
gcloud run services list
gcloud sql instances list
gcloud artifacts repositories list
35. Beginner administrator skills
Know how to:
- select a project,
- enable APIs,
- manage IAM,
- deploy Compute Engine/Cloud Run,
- understand VPC/firewall rules,
- use service accounts,
- inspect logs,
- watch cost.
36. Sources and further learning
Official documentation: https://cloud.google.com/docs
Main idea
GCP is easiest to understand around projects, service accounts, VPC and managed platforms such as Cloud Run. Start simple before reaching for GKE.