1: Designing and planning a cloud solution architecture

Learning Objectives

By the end of this chapter, you will be able to:

• Design a Google Cloud solution infrastructure that satisfies business requirements, including use cases, continuity, cost, integration, data movement, workload disposition, success measurement, security and compliance, and observability.
• Design a Google Cloud solution infrastructure that satisfies technical requirements, including Well-Architected Framework alignment, high availability, flexibility, scalability, performance, Gemini Cloud Assist support, and backup and recovery.
• Select and justify network, storage, and compute resources, including hybrid integration, AI and machine learning services, cloud-native networking constructs, data processing, storage tiers, and compute platform mapping.
• Build a migration plan that integrates with existing systems, uses Migration Center, applies migration methodologies, and accounts for licensing and financial impact.
• Plan for the evolution of the architecture through cloud and technology improvements, changing business needs, and a cloud-first design approach.

Executive Summary

• Architecture decisions are not technology choices first; they are business-requirement decisions that get expressed in Google Cloud services through documented trade-offs and decision anchors.
• The Google Cloud Well-Architected Framework is the controlling reference for any technical design question on the exam: operational excellence, security and privacy and compliance, reliability, cost optimization, performance optimization, and sustainability.
• Resource selection follows a strict mapping: workload shape determines compute, data shape determines storage, traffic shape determines networking, and AI workload type determines whether Vertex AI, Model Garden, Agent Builder, or AI Hypercomputer is appropriate.
• Migration design is a planning artifact, not a tool run. Google Cloud Migration Center is the assessment and wave-planning surface that supplies the inputs for the architectural diagrams, dependency graphs, and license calculations the exam expects you to produce.
• A production-grade architecture includes its own evolution model: cloud-first defaults, scheduled re-assessment, and a backlog of improvements driven by Gemini Cloud Assist insights, FinOps reviews, and changing regulatory or business posture.

Assumptions

• The reader has practitioner experience with Google Cloud and is preparing for an architect-tier scenario exam, not a foundational exam.
• Regional examples use representative regions such as us-central1, europe-west1, and asia-southeast1; specific products are available in their published region lists and the exam does not require memorizing region-by-region availability.
• All references to limits and quotas reflect documented Google Cloud platform values; any organization-specific overrides are noted explicitly.
• Terminology uses the official Google Cloud product names (for example, Cloud Run, Cloud Run functions, BigQuery, Cloud Storage, Spanner, Bigtable, Pub/Sub, Cloud Composer, Looker, Vertex AI, Agent Builder, Model Garden, AI Hypercomputer).