Your organization requires a highly available and scalable web application with session affinity requirements. What architecture best supports these requirements?

Correct answer: App Service plan with multiple instances and Azure Cache for Redis for session management

Option C is correct because an App Service plan with multiple instances satisfies the high availability and scalability requirements, while Azure Cache for Redis provides a distributed, low-latency session store that preserves session affinity across all instances without binding users to a single node. Option A is incorrect because a single App Service instance with sticky sessions is neither highly available nor scalable; a single instance is a single point of failure and cannot scale horizontally while maintaining true session affinity. Option B is incorrect because Azure Cosmos DB is a globally distributed database service and is not designed as a session cache; introducing it for session state management adds unnecessary latency and cost compared to Redis. Option D is incorrect because Azure Virtual Machines with internal load balancing require significantly more operational overhead for OS patching and scaling, and shared storage for sessions introduces latency and complexity not suited to a managed web application architecture.