Microsoft Azure Solutions Architect Expert AZ-305 Practice Question

Azure SQL Database elastic pools let you place many individual databases into a single pool that shares a set of vCores and memory. When one or more databases experience a workload spike, they can consume more of the pool's available compute, while idle databases consume very little, keeping overall costs low. Microsoft specifically recommends elastic pools for multi-tenant SaaS scenarios that involve hundreds or thousands of small databases with unpredictable usage patterns.

By contrast:

• Azure SQL Database in the Hyperscale tier provisions compute per database, so each tenant requires its own dedicated resources, increasing costs linearly.
• An Azure SQL Managed Instance provisions compute at the instance level; although databases inside the same instance share resources, scaling to thousands of tenant databases would likely require multiple instances, each incurring its own fixed compute cost, reducing cost efficiency for mostly idle workloads.
• Deploying thousands of separate serverless single databases would eliminate cross-database resource sharing; although each database can auto-pause, you still pay per database and cannot pool compute across tenants.

Therefore, elastic pools best satisfy the requirement for automatic, cost-efficient redistribution of compute across many small, variable-load databases.