- Domain 2 Overview: What Microsoft Actually Tests
- Storage Accounts and Access Tiers
- Choosing Between Relational and Non-Relational Databases
- Data Integration and Migration Design
- How AZ-305 Tests Data Storage Decisions
- Scheduling Domain 2 Inside Your Study Plan
- Common Mistakes Candidates Make in This Domain
- Who Hires for These Skills
- Frequently Asked Questions
- Domain 2 covers 20-25% of AZ-305, roughly one in five scored questions.
- Focus on choosing storage tiers, replication, and database services for given business scenarios.
- You must justify tradeoffs between Cosmos DB, Azure SQL, and Synapse, not just define them.
- Case studies in this domain often combine data storage with cost and compliance constraints.
Domain 2 Overview: What Microsoft Actually Tests
Design data storage solutions sits at 20-25% of the AZ-305 blueprint, making it the second-largest of the four domains behind Domain 4: Design infrastructure solutions. That weighting means somewhere between a fifth and a quarter of your scored questions will hinge on how well you can architect storage and database solutions for realistic business scenarios, not just recite service names.
Unlike a hands-on data engineering exam, AZ-305 does not ask you to write Azure SQL queries or configure a Cosmos DB container line by line. It asks you to read a scenario, identify constraints such as latency, consistency, cost, or compliance, and then pick the storage architecture that satisfies all of them simultaneously. This is the core skill Microsoft is testing across the entire exam, and it's especially visible in Domain 2 because storage decisions almost always involve tradeoffs.
Storage Accounts and Access Tiers
A significant chunk of Domain 2 revolves around Azure Storage accounts: blob storage, file shares, and the access tiers (hot, cool, cold, and archive) that control cost and retrieval latency. You need to be comfortable recommending redundancy options - LRS, ZRS, GRS, and GZRS - based on stated durability and disaster recovery requirements, and explaining how lifecycle management policies automate tier transitions to reduce storage spend over time.
Storage Account Design
Candidates must map business requirements to concrete storage configuration choices, not just describe what each option does.
- Match access tiers to data access frequency and retrieval SLAs
- Select redundancy (LRS/ZRS/GRS/GZRS) based on regional failure tolerance
- Apply lifecycle management to move aging data to cool or archive tiers
- Recommend security controls: private endpoints, shared access signatures, encryption scope
Expect scenario language like "the solution must minimize cost for data accessed less than once a month while supporting retrieval within hours." That phrasing is a direct pointer toward cool or archive tiers combined with lifecycle rules, and recognizing these patterns quickly is what separates a fast, confident test-taker from someone re-reading the question three times.
Choosing Between Relational and Non-Relational Databases
The heart of Domain 2 is database selection. Microsoft expects you to differentiate Azure SQL Database, Azure SQL Managed Instance, Azure Database for PostgreSQL/MySQL, Azure Cosmos DB, and Azure Synapse Analytics based on workload characteristics rather than marketing descriptions.
- Relational vs. non-relational: Know when a scenario's data model (structured, transactional, foreign-key relationships) points to a relational engine versus when flexible schema and horizontal scale point to Cosmos DB.
- Consistency models: Cosmos DB's five consistency levels (strong, bounded staleness, session, consistent prefix, eventual) show up as scenario-driven questions where you must pick the level that balances latency against correctness.
- Compute tiers and scaling: Distinguish DTU-based versus vCore-based purchasing models for Azure SQL, and know when elastic pools make sense for multi-tenant workloads.
- Analytics workloads: Recognize when a scenario calls for Synapse Analytics or a dedicated SQL pool versus a transactional database engine.
Key Takeaway
When a question mentions "global distribution" and "low-latency writes," think Cosmos DB. When it mentions "complex joins" or "existing SQL Server workload," think Azure SQL Database or Managed Instance. Train yourself to spot these trigger phrases.
| Service | Best Fit Scenario | Key Design Concern |
|---|---|---|
| Azure SQL Database | Modern cloud-native relational apps | Purchasing model, elastic pools, geo-replication |
| Azure SQL Managed Instance | Lift-and-shift SQL Server migrations | Instance-level features, VNet integration |
| Azure Cosmos DB | Global, multi-region, flexible schema apps | Consistency level vs. latency tradeoff |
| Azure Synapse Analytics | Large-scale analytics and data warehousing | Dedicated vs. serverless SQL pools |
| Azure Database for PostgreSQL/MySQL | Open-source relational migrations | Flexible server vs. single server tiers |
Data Integration and Migration Design
Domain 2 also covers how data moves into and across Azure. You should be able to recommend Azure Data Factory or Azure Synapse pipelines for orchestration, and choose between Azure Database Migration Service, backup/restore, or third-party tooling depending on downtime tolerance and source database type.
Migration and Integration Topics
Expect scenarios where a company is moving on-premises databases to Azure or integrating multiple data sources.
- Online vs. offline migration strategies and their downtime implications
- Data Factory pipeline design for ETL/ELT scenarios
- Hybrid connectivity requirements (VPN, ExpressRoute) for ongoing data sync
- Backup, point-in-time restore, and geo-restore for business continuity of data
Notice the overlap here with Domain 3: Design business continuity solutions - backup and restore design questions can appear in either domain depending on framing, which is a good reminder that the four domains are not siloed in the real exam.
How AZ-305 Tests Data Storage Decisions
AZ-305 uses a mix of formats - multiple choice, drag-and-drop, hot area, and case studies - and Domain 2 content shows up across all of them. A typical case study will describe a company's existing data estate, list explicit technical and business requirements, and then ask several questions about the storage tier, database service, or migration approach that satisfies those requirements. Because Microsoft Learn access is available during the exam within its own domain while the clock keeps running, you can look up service capabilities, but you cannot look up which service fits a specific business tradeoff - that judgment has to be yours.
If you're still building a mental model of how all four domains interact, the AZ-305 Exam Domains 2026: Complete Guide to All 4 Content Areas breaks down each domain's weighting and scope side by side, which helps you see where Domain 2 topics bleed into the others.
Scheduling Domain 2 Inside Your Study Plan
Because Domain 2 carries a 20-25% weight, it deserves a dedicated block in your study calendar rather than being folded into general "Azure services" review. A practical approach is to study it right after governance and identity so you already understand role-based access control and networking basics that apply to securing storage accounts and databases.
Storage Fundamentals
- Storage account tiers, redundancy options, and lifecycle policies
- Blob vs. file vs. table vs. queue storage use cases
Database Selection
- Relational vs. non-relational decision criteria
- Cosmos DB consistency levels and partitioning strategy
Migration and Integration
- Data Factory pipeline scenarios
- Migration tooling based on downtime tolerance
Practice and Review
- Full-length case studies mixing storage with governance topics
- Timed practice questions on ../ to simulate exam pacing
This kind of structured sequencing is one piece of a larger plan - for the complete week-by-week roadmap across all four domains, see the AZ-305 Study Guide 2026: How to Pass on Your First Attempt.
Common Mistakes Candidates Make in This Domain
Most candidates don't fail Domain 2 because they don't know the services - they fail because they misread the constraint that matters most in a given scenario. A few recurring patterns:
- Defaulting to Cosmos DB for every "scalable" requirement: Scalability alone doesn't rule out Azure SQL with read replicas or elastic pools.
- Ignoring cost constraints in tiering questions: A scenario asking for "lowest cost with acceptable retrieval delay" is pointing at archive tier, not cool tier.
- Confusing backup features with high availability features: Geo-redundant storage protects against regional failure; it is not a substitute for active-active database replication.
- Skipping consistency-level nuance in Cosmos DB questions: Many test-takers know Cosmos DB exists but can't distinguish session consistency from bounded staleness under pressure.
If you want a broader sense of where candidates typically lose points across the whole exam, the How Hard Is the AZ-305 Exam? Complete Difficulty Guide 2026 article and the AZ-305 Pass Rate 2026: What the Data Shows piece both address this from different angles.
Who Hires for These Skills
Employers hiring for cloud architect, data platform architect, and infrastructure consultant roles look for exactly the judgment Domain 2 tests: the ability to select and justify a storage or database architecture rather than just deploy one. This domain's content maps closely to day-to-day responsibilities in organizations migrating on-premises SQL Server estates, building global-scale applications, or consolidating data platforms across business units.
Because the AZ-305 Solutions Architect Expert credential requires candidates to already hold Azure Administrator Associate, employers treat it as a signal of layered, verified experience rather than an entry-level checkbox. For a deeper look at how this shows up in job postings and compensation, see the AZ-305 Jobs overview and the AZ-305 Salary Guide 2026: Complete Earnings Analysis.
Key Takeaway
Domain 2 mastery is directly transferable to real architecture reviews - practicing it well pays off beyond the exam itself.
To sanity-check your readiness before exam day, run timed practice sets on the AZ-305 practice test platform that specifically isolate storage and database scenarios, then review every miss against the exact requirement you overlooked. Repeating this cycle a few times on practice questions is often more effective than re-reading documentation.
Frequently Asked Questions
Microsoft weights Domain 2 at 20-25% of the exam. Since most Microsoft certification exams typically contain 40-60 questions, that translates to roughly 8-15 questions touching storage and database design, though the exact count varies by exam version.
You don't need to write queries, but you do need practical familiarity with configuration options like consistency levels, purchasing models, and redundancy settings, since questions test applied judgment, not definitions.
It's not inherently harder, but its scenario-heavy questions require juggling multiple constraints at once. Candidates who study services in isolation often struggle more here than in domains with more straightforward configuration questions.
Data storage decisions often depend on networking and compute choices covered in Domain 4: Design infrastructure solutions, such as private endpoints or VM-hosted databases, so the domains are best studied with awareness of their overlap.
Start with What Is AZ-305? and AZ-305 Certification for foundational context, then move into domain-specific study guides once you understand the exam's overall structure.
- AZ-305 Domain 1: Design identity, governance, and monitoring solutions (25-30%) - Complete Study Guide 2026
- AZ-305 Domain 3: Design business continuity solutions (15-20%) - Complete Study Guide 2026
- AZ-305 Domain 4: Design infrastructure solutions (30-35%) - Complete Study Guide 2026
- AZ-305 Exam Domains 2026: Complete Guide to All 4 Content Areas