Why GCP?
Data & ML Focus:
- 11% market share (third largest)
- Strength in big data, analytics, ML
- Used by Google (obviously) and data-heavy companies
- Best-in-class machine learning tools
- Excellent for data science teams
- Strong Kubernetes ecosystem (Google created K8s)
GCP Global Infrastructure
Regions
40+ regions worldwide
Zones
Each region has 3+ zones (like availability zones)
# List zones
gcloud compute zones list
# Set default zone
gcloud config set compute/zone us-central1-aGCP Terminology
Project: Container for all resources and billing Region: Geographic location Zone: Data center within region Service: Individual cloud service (Compute Engine, Cloud Storage, etc.)
# Create project
gcloud projects create my-project
# List projects
gcloud projects list
# Set current project
gcloud config set project my-projectCore GCP Services
Compute
Compute Engine - VMs (like EC2)
gcloud compute instances create my-instance --image=ubuntu-2004-lts --zone=us-central1-aApp Engine - PaaS (like Elastic Beanstalk)
gcloud app deployCloud Functions - Serverless (like Lambda)
gcloud functions deploy myFunction --runtime python39 --trigger-httpGKE (Google Kubernetes Engine) - Managed Kubernetes
gcloud container clusters create my-cluster --zone us-central1-aStorage
Cloud Storage - Object storage (like S3)
gcloud storage buckets create gs://my-bucket
gcloud storage cp myfile.txt gs://my-bucket/Persistent Disk - Block storage (like EBS)
gcloud compute disks create my-disk --size=100GB --zone=us-central1-aFirestore/Datastore - Cloud file storage
gcloud firestore databases create --location=us-central1Database
Cloud SQL - Managed MySQL, PostgreSQL, SQL Server
gcloud sql instances create my-database --database-version=POSTGRES_13BigQuery - Data warehouse (analyze terabytes in seconds)
gcloud bigquery datasets create my_datasetFirestore - NoSQL, real-time database
gcloud firestore databases create --location=us-central1Networking
VPC (Virtual Private Cloud) - Your private network
gcloud compute networks create my-network --subnet-mode=customCloud Load Balancing - Global load balancing
gcloud compute backend-services create my-backend-serviceCloud CDN - Content delivery network
# Enable CDN on backend service
gcloud compute backend-services update my-service --enable-cdnBig Data & ML
Dataflow - Batch and stream processing
gcloud dataflow jobs run my-job --gcs-location gs://my-bucket/templatePub/Sub - Message queuing
gcloud pubsub topics create my-topicVertex AI - Machine learning platform
gcloud ai endpoints listUnique GCP Strengths
š BigQuery - Analyze massive datasets instantly š Vertex AI - Complete ML platform (AutoML, custom training) š« Pub/Sub - Global messaging system š Dataflow - Apache Beam as managed service š Cloud Storage - Highly consistent, global availability š Cloud Audit Logs - Comprehensive audit trail
Pricing
Pay-As-You-Go (like AWS/Azure) Committed Use Discounts - Contract pricing Free Tier - $300 credit for 90 days Per-second Billing (not hourly like AWS)
GCP vs AWS vs Azure
| Aspect | GCP | AWS | Azure |
|---|---|---|---|
| Data/ML | Best | Good | Good |
| Kubernetes | Best | Good | Good |
| Big Data | Best | Good | Good |
| Ease of Use | Good | Fair | Good |
| Global Reach | Good | Best | Best |
| Enterprise Features | Good | Best | Best |
| Market Share | 11% | 32% | 23% |