Design Data Storage Solutions
This domain covers selecting the right relational, non-relational, and object storage for a workload's consistency, scale, and cost requirements. Match the data characteristics to the appropriate Azure service.
Relational data design
Choose Azure SQL Database for managed single databases or elastic pools, and SQL Managed Instance when you need near-full SQL Server compatibility for lift-and-shift. Use read replicas and appropriate service tiers (for example the Hyperscale tier for very large databases) to meet performance and scale needs. Open-source options include Azure Database for PostgreSQL and MySQL.
Non-relational and NoSQL design
Azure Cosmos DB serves globally distributed, low-latency NoSQL workloads and lets you tune consistency levels from strong to eventual to balance latency, availability, and correctness. Choose a partition key with high cardinality to distribute throughput and avoid hot partitions. Select the appropriate API (for example NoSQL, MongoDB, or Cassandra) for compatibility.
Object and file storage design
Azure Blob storage stores unstructured data with hot, cool, cold, and archive access tiers and lifecycle management to move data automatically as it ages. Azure Data Lake Storage Gen2 adds a hierarchical namespace for analytics. Azure Files and Azure NetApp Files provide managed file shares for lift-and-shift and high-performance scenarios.
Data protection and movement
Encrypt data at rest with platform-managed or customer-managed keys in Azure Key Vault, and secure access with private endpoints and network rules. Design data ingestion and movement with Azure Data Factory or Azure Synapse pipelines, and choose redundancy (LRS, ZRS, GRS) to meet durability and residency requirements.