The previous chapters dealt with local and client-server ObjectGrid interaction. Local ObjectGrid instances run in the same memory process as the business application. Access to objects stored in the grid is extremely fast, and there are no network hops or routing done on ObjectGrid operations. The disadvantage with a local ObjectGrid instance is that all objects stored in the grid must fit into the heap space of one JVM.
The client-server distributed ObjectGrid instances we studied in the previous chapters overcame that single heap space disadvantage by combining the resources of multiple JVMs on multiple servers. These combined resources hide behind the façade of an ObjectGrid instance. The ObjectGrid instance has far more CPU, memory, and network I/O available to it than the resources available to any single client.
In this chapter, we'll learn how to use those resources held by the ObjectGrid instance to co-locate data and business logic on a single JVM. The...