AZ-305 Design business continuity solutions Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Deploy AKS in two regions with Azure Traffic Manager. Use Azure Cosmos DB with multi-region writes. Use Azure Cache for Redis Enterprise with active geo-replication.

  Why it's wrong here

  Deploying AKS with Azure Traffic Manager for failover is problematic for a 30-second RTO. Traffic Manager relies on DNS propagation, which can introduce delays exceeding this strict requirement during a regional outage, impacting the application's availability. While active geo-replication for Azure Cache for Redis Enterprise meets RPO 0 and low RTO for caching, the AKS component fails. This option is tempting as Traffic Manager is a standard solution for multi-region traffic distribution and failover in less stringent RTO scenarios.

• ✗

  Deploy AKS in two regions with Azure Front Door. Use Azure Cosmos DB with a single write region and auto-failover. Use Azure Cache for Redis Standard with geo-replication.

  Why it's wrong here

  Using a single write region in Azure Cosmos DB with automatic failover does not guarantee zero data loss (RPO=0) because replicated writes to the secondary region are asynchronous depending on consistency level, and if the primary fails before replication, data can be lost. Additionally, Azure Cache for Redis Standard tier does not support geo-replication at all, so the stated architecture cannot provide any cross-region cache redundancy. These two gaps make the solution unable to meet a 30-second RTO and RPO=0 requirement.

• ✓

  Deploy AKS in two regions with Azure Front Door. Use Azure Cosmos DB with multi-region writes. Use Azure Cache for Redis Enterprise with active geo-replication.

  Why this is correct

  By using Azure Cosmos DB multi-region writes, every write is acknowledged in all configured regions, guaranteeing zero data loss (RPO) and enabling immediate failover. Azure Front Door sits in front of AKS clusters in two regions, performing health-probe-based endpoint selection and traffic redirection within seconds, which supports the strict 30-second RTO. Azure Cache for Redis Enterprise with active geo-replication creates active-active caches that allow simultaneous reads and writes in both regions without data loss. This combination provides a truly active-active data plane and global HTTP routing, making it the only architecture that fully satisfies all requirements.

• ✗

  Use Azure SQL Database with auto-failover groups for the database tier. Deploy AKS in two regions with Azure Front Door. Use Azure Cache for Redis Enterprise with active geo-replication.

  Why it's wrong here

  This option fails because Azure SQL Database auto-failover groups rely on a single primary that replicates asynchronously to a secondary region; even though failover is automatic, the potentially lagging transaction log replica means committed transactions can be lost, so RPO=0 is not achievable. The AKS and Redis Enterprise layers are well-designed, but the database tier remains active-passive and cannot match the zero-loss guarantee provided by Cosmos DB multi-region writes. Therefore, this choice does not meet the application's strict RPO and RTO requirements even though it appears to offer a managed database failover.

Go deeper

Related to this question