Hands-On: System Design for GCP

Designing scalable systems on Google Cloud Platform requires knowing exactly which managed services to choose and how to securely connect them. This course is a hands-on walkthrough of building a complete, production-ready architecture on GCP. You will start at the edge, configuring Cloud Load Balancing, edge caching with Cloud CDN, and API ingress with API Gateway to route traffic to compute workloads running side by side on Cloud Run and GKE. With the entrypoint established, you will design the data layer. You will connect Cloud Run to Cloud SQL, scale reads using replicas, implement a cache-aside pattern with Memorystore, and handle global scale using Cloud Spanner. For non-relational data, you will store documents in Firestore, manage high-throughput writes with Bigtable, and serve assets directly from Cloud Storage using signed-URL direct uploads.

From there, the course moves into asynchronous messaging and event-driven design. You will decouple services using Pub/Sub for fan-out and dead-letter queues, manage task execution with Cloud Tasks, and build an analytics pipeline by streaming and batching data into BigQuery to serve CQRS read models. A production system must also be secure and observable, so you will configure public and private subnets, isolate the data layer, and deploy Cloud Armor at the edge. You will integrate authentication via Identity Platform, lock down the data path, and manage configuration with Secret Manager. Finally, you will ensure redundancy across availability zones and set up metrics, logs, traces, and alerting using Cloud Monitoring. Through a series of capstone projects in every module, you will tie these concepts together, culminating in a secure, segmented, and fully observable multi-tier application.