Overview of use cases where (clustered) cache/memory-grid could be used:

• throttling
• cache (static) external content (e.g. codebases, static file content, ...)
• cache internal OpenHub data
  • cache configuration properties
  • use shared memory for saving runtime data, e.g. node evidence or evidence of running scheduled jobs
• cache for specific business logic implementations on projects

We had the following requirements when we have been selecting suitable cache implementation:

• uses Spring cache abstraction layer
• supports Spring Session (http://projects.spring.io/spring-session/)
• cache/memory-grid should support both modes by changing configuration only (without needs to change application code) - single application or cluster. In other words cluster should be option.
• first-class support for Spring Boot (auto-configuration)
• support in Apache Camel

We selected Hazelcast as default implementation.