Performance and Efficiency Issues in Distributed Enterprise Management

In an attempt to improve overall performance in distributed enterprise management architectures we propose a caching model and the related interaction semantics and update policies. Our work is motivated by the increasing need for scalable, distributed architectures for integrated management within the enterprise network scale. The integration of network, system and application management and the wide area deployment of management functionality imposes heavy performance requirements, in direct conflict with the increased volume of management traffic produced. We construct a hierarchical caching scheme that attempts to aid distributed management architectures and identify the roles of management components and their interaction semantics. In order to evaluate the proposed model we employ both analytical tools and simulations (experiments with real data traces) that provide insight on expected performance gain.