An organization is designing a high-availability solution for a mission-critical database workload in Microsoft Azure. The solution must protect the workload against entire datacenter failures while supporting synchronous data replication requiring an inter-site round-trip network latency of less than approximately 2 milliseconds. Which deployment strategy meets these architectural and latency requirements?

Deploying the database across multiple Availability Zones within the same Azure region: ### What Availability Zones Are **Availability Zones** are physically separate locations made up of one or more datacenters within an **Azure region**. Each zone is equipped with independent power, cooling, and networking infrastructure. This physical isolation ensures that if an issue arises in one datacenter facility, the remaining zones in the region continue to operate without disruption. ### How It Meets the Requirements - **Datacenter-wide protection**: Because each Availability Zone represents distinct physical datacenters with independent facilities, distributing database nodes across zones safeguards the application against full datacenter outages. - **Low-latency connectivity**: Availability Zones within a region are connected via a high-performance, private fiber-optic network designed to achieve **round-trip latency of less than approximately 2 milliseconds**, making it ideal for **synchronous replication**. ### Key Advantages and Capabilities - **Resiliency against facility failures**: Protects applications from localized disruptions such as power loss, cooling failure, or physical infrastructure damage. - **High-throughput data paths**: Inter-zone connectivity uses dedicated, ultra-low-latency networking infrastructure optimized for real-time data synchronization. - **Sequenced platform updates**: Azure rolls out service and platform maintenance to one zone at a time to keep multi-zone architectures accessible during update cycles. - **Granular high availability**: Allows organizations to achieve high availability SLAs without the latency overhead of routing traffic across distant geographical regions. ### Why This Is Optimal Deploying across **Availability Zones** strikes the ideal balance between failure isolation and low network latency. It eliminates single points of datacenter failure while staying well within the ultra-low latency threshold necessary for synchronous database clustering.