When provisioning an Azure Cosmos DB account, which feature provides redundancy within an Azure region?

Availability Zones (AZs) provide high availability and redundancy within a single Azure region. When you enable this feature for an Azure Cosmos DB account, the service replicates your data across multiple, physically separate datacenters (zones) within that region. This design protects your database from datacenter-level failures, such as power or network outages affecting a single zone. If one zone becomes unavailable, Azure Cosmos DB transparently fails over to a replica in another zone within the same region, ensuring continuous operation without data loss and maintaining the service level agreement (SLA).

A. multi-master replication: This feature provides redundancy and low-latency writes across multiple Azure regions (geo-redundancy), not within a single region.
C. automatic failover: This is a mechanism for disaster recovery that shifts operations from a primary region to a secondary region during a regional outage, addressing inter-region, not intra-region, redundancy.
D. the strong consistency level: This is a data consistency model that guarantees reads receive the most recent committed write. It does not, by itself, provide infrastructure redundancy.

1. Microsoft Learn, "High availability with Azure Cosmos DB": Under the section "High availability within a region," it states, "For high availability within a region, when you provision an Azure Cosmos DB account, you can configure it to use availability zones. Availability zones use physically separate groups of datacenters within a region to provide a high degree of availability for the data in the account."

2. Microsoft Learn, "Configure multi-region writes in your applications that use Azure Cosmos DB": This document describes multi-region writes (formerly multi-master) as a feature to "distribute your data globally" and "associate any number of Azure regions with your Azure Cosmos DB account." This confirms its purpose is for inter-region distribution.

3. Microsoft Learn, "Automatic failover for Azure Cosmos DB": The documentation states, "With automatic failover enabled, in the event of a regional disaster, Azure Cosmos DB will failover your account to the region with the highest failover priority with no user action." This clearly defines it as a regional disaster recovery feature.

4. Microsoft Learn, "Consistency levels in Azure Cosmos DB": This page details the guarantees of different consistency models, such as Strong, Bounded Staleness, and Eventual. The focus is on data correctness and latency trade-offs, not infrastructure resilience against failures.