Large Scale Network-Centric Distributed Systems

5

Data Distribution Management

Azzedine Boukerche and Yunfeng Gu

Contents

5.1 Addressing DDM in Different Network Environments

5.2 DDM in P2P Overlay Networks

5.2.1 Background

5.2.2 Data Space Partitioning and Mapping

5.2.3 Corresponding Overlay Network Support

5.3 DDM in Cluster-Based Network Environments

5.3.1 Basic Concepts in DDM

5.3.1.1 Routing Space

5.3.1.2 Objects, Federates, and Federation

5.3.1.3 Behavior of Entities

5.3.1.4 Other Important Concepts

5.3.1.5 Performance of DDM Implementations

5.3.2 Background

5.3.3 Data Distribution Management Schemes

5.3.3.1 Region-Based DDM Approach

5.3.3.2 Grid-Based DDM Approaches

5.3.3.3 Other DDM Schemes

References

In general, DDM (data distribution management) deals with two basic problems: how to distribute data generated at the application layer among underlying nodes in the distributed system and how to retrieve data whenever it is necessary. Alternatively, these two problems can also be described as partitioning of the application data space and mapping from the application data space into the system identifier space. The concept of DDM can also be found in HLA/RTI (high-level architecture/run-time infrastructure) under IEEE Standard 1516. It is one of seven service APIs provided by RTI. HLA is a general-purpose architecture for real-time distributed interactive simulation systems. It is used to support the portability and interoperability of a large number of distributed simulations across various computing platforms. RTI ...