Google Cloud Fundamentals Course
A structured beginner course covering Google Cloud concepts, core services, data, networking, security, and pricing.
What you will learn
- Explain Google Cloud concepts and global infrastructure
- Identify core compute, storage, and networking services
- Describe Google Cloud data and machine learning services
- Apply Google Cloud security, identity, and governance practices
- Understand pricing, billing, and operational management
Before you start
- No prior Google Cloud experience required
- Basic computer and internet familiarity
- A free Google Cloud account is useful but optional
Lesson 1 Google Cloud Concepts and Global Infrastructure
Google Cloud organizes resources through a hierarchy: organization, folders, projects, and resources. Projects are the basic unit where services run, and resources within a project share billing and access policies.
Google Cloud operates in Regions and Zones. A Region is a geographic area with multiple Zones, and each Zone contains one or more data centers. Choosing a Region close to users reduces latency, while spreading workloads across Zones improves availability.
Cloud benefits include scalability, elasticity, pay-as-you-go pricing, high availability, and reduced infrastructure management. Customers use APIs and console tools to provision resources on demand.
Hierarchy drill: draw the hierarchy organization -> folders -> projects -> resources, and label what each level controls. A project isolates billing and IAM; folders group projects; the organization sets top-level policies.
Example
A company creates a project for its web application, deploys Compute Engine instances in two Zones, and uses Cloud Load Balancing to route traffic. If one Zone fails, the application continues to serve users from the other Zone.
Worked example: A company has finance and engineering teams.
Create two folders, place one project in each, and apply different IAM policies at the folder level.
Lesson 2 Core Compute, Storage, and Networking Services
Compute Engine provides virtual machines with flexible machine types. Google Kubernetes Engine (GKE) runs containerized workloads, Cloud Run deploys containers as serverless services, and Cloud Functions runs code in response to events. App Engine provides a managed platform for web applications.
Cloud Storage stores objects such as images and backups with different storage classes. Persistent Disk offers durable block storage for VMs, and Filestore provides managed network file storage.
Virtual Private Cloud (VPC) creates isolated networks with subnets and firewall rules. Cloud Load Balancing distributes traffic globally, Cloud CDN caches content at edge locations, and Cloud DNS translates domain names to IP addresses.
Service selection: choose Compute Engine for full VM control, GKE for containers at scale, Cloud Run for serverless containers, Cloud Functions for event-driven code, and App Engine for managed web apps. Match the workload before memorizing features.
Example
A startup builds a mobile API with Cloud Run, stores user uploads in Cloud Storage, and places Cloud CDN in front of static assets. This removes server management while keeping the application scalable.
Worked example: A team wants to run a containerized web service without managing servers.
Cloud Run scales automatically and charges only for requests and memory during execution.
Lesson 3 Data, Analytics, and Machine Learning Services
BigQuery is a serverless data warehouse for analyzing large datasets with SQL. Cloud SQL provides managed relational databases, Cloud Spanner offers globally distributed relational storage, and Firestore is a flexible NoSQL document database. Bigtable handles high-throughput NoSQL workloads.
Pub/Sub connects producers and consumers through asynchronous messaging. Dataflow processes streaming and batch data pipelines, while Dataproc runs managed Spark and Hadoop clusters.
Vertex AI provides tools to build and deploy machine learning models. BigQuery ML creates models with SQL, AutoML trains models without writing code, and Cloud Translation, Vision, and Speech APIs add ready-made AI capabilities.
Data service mapping: BigQuery for SQL analytics over huge datasets, Cloud SQL for managed relational databases, Cloud Spanner for global consistency, Firestore for flexible NoSQL documents, Bigtable for high-throughput NoSQL, and Pub/Sub for async messaging.
Example
An e-commerce team streams user events into Pub/Sub, processes them with Dataflow, and analyzes the results in BigQuery. They then use BigQuery ML to predict customer segments without moving data to a separate platform.
Worked example: A company needs real-time event ingestion and later SQL analysis.
Pub/Sub receives events, and BigQuery stores and analyzes the processed data.
Lesson 4 Security, Identity, and Governance
Identity and Access Management (IAM) controls who can access Google Cloud resources. Roles grant permissions, and service accounts provide identities for applications. Follow the principle of least privilege and enable multi-factor authentication for users.
VPC firewall rules filter network traffic, Cloud Armor protects applications from attacks, and Security Command Center provides a central view of security risks. Cloud KMS manages encryption keys, and Secret Manager stores sensitive values such as passwords and API keys.
Google Cloud supports compliance standards including ISO, SOC, and GDPR. Organization policies enforce rules across projects, and data residency options help meet regional requirements.
Least privilege drill: give each user or service only the roles needed for its task, use service accounts for applications, and review permissions regularly. Enable MFA, use VPC firewall rules to filter traffic, and monitor with Cloud Logging and Security Command Center.
Example
A finance application stores database credentials in Secret Manager, uses Cloud KMS to encrypt data, and applies an organization policy that blocks public Cloud Storage buckets. Security Command Center monitors the environment for misconfigurations.
Worked example: A batch job only needs to read one bucket.
Create a service account with a single storage viewer role for that bucket instead of owner access.
Lesson 5 Pricing, Billing, and Operations
Google Cloud pricing depends on the service, machine type, region, and usage duration. The Pricing Calculator estimates monthly costs, and budgets with alerts notify teams before spending exceeds limits.
Sustained use discounts and committed use discounts reduce cost for predictable workloads. Preemptible and Spot VMs provide lower prices for interruption-tolerant jobs. Labels organize resources for cost reporting.
Cloud Monitoring tracks metrics and uptime, Cloud Logging stores logs, and Cloud Error Reporting surfaces application errors. Service health pages communicate outages, and quotas limit resource usage to protect the shared infrastructure.
Cost control loop: estimate with Pricing Calculator, set budgets with alerts, group resources with labels, use sustained-use and committed-use discounts for predictable work, and choose preemptible VMs for fault-tolerant tasks. Monitor with Cloud Monitoring.
Example
A batch analytics team uses committed use discounts for steady BigQuery slots, runs short processing jobs on preemptible VMs, and sets a monthly budget alert at 80 percent. Cloud Monitoring sends an alert if job failure rates increase.
Worked example: A nightly data-processing job can be interrupted.
Preemptible VMs reduce cost; the job must be designed to retry and resume.