AZ-305 Designing Microsoft Azure Infrastructure Solutions Exam

Recommend a Database Service Tier and Compute Tier

Map Workload Requirements to Service and Compute Offerings

Mapping your workload requirements to the correct service tier and compute configuration involves evaluating transaction volume, concurrency, and I/O throughput. You also consider availability requirements and cost constraints to choose between the vCore and DTU purchasing models. In the vCore model, you select the number of vCores, the amount of memory, and storage size independently. In the DTU model, you choose a bundled set of compute, memory, and I/O resources. This approach ensures SLA compliance while optimizing for cost efficiency.

• Transaction volume and concurrency drive the need for more compute resources (vCores or DTUs).
• I/O throughput and latency requirements influence the choice of storage architecture.
• Growth and scale determine whether flexible scaling in Hyperscale is required.
• Budget targets and commitment options guide the selection of reserved instances or hybrid benefits.

General Purpose Tier

General Purpose service tier uses a remote storage model with separated compute and storage layers. It’s ideal for budget-oriented workloads that require balanced compute, moderate I/O performance, and high availability through Azure Blob storage. Typical scenarios include moderate transaction rates and no strict low-latency demands. You can choose between provisioned or serverless compute tiers to pay for fixed or usage-based resources, respectively. The flexibility helps maintain cost efficiency for general business applications.

Business Critical Tier

The Business Critical tier delivers consistently low I/O latency (1–2 ms on average) by using local SSD storage and multiple high availability replicas. It’s suited for OLTP systems with high transaction volumes, demanding fast failover, and built-in read scale-out for reporting on a free secondary replica. Advanced features such as automatic page repair and zone-redundant availability further protect against corruption and zonal failures. This tier has a higher cost but offers a higher SLA and resilience for mission-critical applications. Choose appropriate vCore sizes within this tier to match peak concurrency and throughput needs.

Hyperscale Tier

The Hyperscale tier provides a cloud-native architecture with independently scalable compute and storage, supporting up to 128 TB of data. You can dynamically adjust vCores without waiting for data copy, and scale out read workloads with named replicas. This tier is ideal for both enormous data volumes and mixed OLTP/analytical or HTAP workloads that outgrow traditional limits. Billing is based on actual storage usage and provisioned compute, allowing efficient cost management. Hyperscale’s flexible model ensures rapid scaling, fast backup/restore, and high availability for the broadest range of database scenarios.

Conclusion

In summary, selecting the appropriate service tier and compute tier involves understanding your workload patterns, including transaction volume, concurrency, and I/O throughput. The General Purpose tier suits budget-sensitive applications with balanced resource needs. The Business Critical tier is designed for high-performance OLTP systems requiring high availability and low latency. Finally, the Hyperscale tier addresses massive data volumes with advanced scalability and flexibility. Evaluating these factors helps ensure optimal SLA compliance and cost efficiency for your Azure SQL database solutions.