JavaSpaces™ Principles, Patterns, and Practice

3.1. Introduction to Distributed Data Structures

Space-based applications are typically designed around distributed data structures. Conceptually a distributed data structure is a data structure that can be accessed and manipulated by multiple processes at the same time. In most distributed computing models, distributed data structures are hard to achieve. Message passing and remote method invocation systems provide a good example of the difficulty. As Figure 3.1 illustrates, these systems tend to barricade data structures behind one central manager process, and processes that want to perform work on the data structure must “wait in line” to ask the manager process to access or alter a piece of data on their behalf. Attempts to parallelize or ...