Which components use cluster shared memory in a Mule cluster?

• A. VM queues and Object Stores
• B. HTTP endpoints
• C. Anypoint Connectors
• D. External Caches

Answer: (A)

In a Mule cluster, components that need to share state quickly across all cluster members rely on cluster shared memory. VM queues and the Object Store are designed to hold in-memory state and be accessible by any node in the cluster, which enables fast failover and coordinated processing. VM queues store messages in memory and must remain usable if the active node changes, while the Object Store keeps state and data needed by various apps and services across the cluster. This shared memory approach ensures that stateful interactions can continue smoothly without losing data when a node fails.

HTTP endpoints are typically stateless, so they don’t require or depend on cluster-shared memory. Anypoint Connectors and External Caches aren’t inherently using cluster shared memory by default—connectors may use the Object Store for specific state, and external caches are separate caching mechanisms, but they aren’t what’s meant by cluster-shared memory in this context.