You are designing a globally distributed order processing system on Azure Cosmos DB for the Core (SQL) API. The system must accept writes in both East US and West Europe with no data loss during a single-region outage and fail over automatically. Read latency should be minimized for local users. What should you configure?

To meet the requirements, enabling multi-region writes in Azure Cosmos DB for the Core (SQL) API allows both East US and West Europe to accept writes, preventing data loss during a single-region outage. Configuring automatic failover with a region priority list ensures the system automatically switches write regions if one becomes unavailable. Using session consistency for clients in each region provides a good balance between read performance and data consistency, guaranteeing that a client will always read its own writes, which minimizes read latency for local users. Using a single write region with strong consistency would not allow writes in both regions simultaneously. Enabling multiple write regions with strong consistency would introduce high latency due to the need for global consensus on every write, making it unsuitable for minimizing read latency. Azure Table storage lacks the globally distributed, multi-master write capabilities and automatic failover required. Using MongoDB on VMs introduces significant management overhead and doesn't leverage Cosmos DB's native capabilities.