Realizing Network Control Policies Using Distributed Action Plans

  • Authors:
  • Madhur Kohli;Jorge Lobo

  • Affiliations:
  • Bell Labs, 600 Mountain Avenue, Murray Hill, New Jersey 07974/ madhur@research.bell-labs.com;Teltier Technologies, 60 Walnut Avenue, Clark, New Jersey 07066/ jlobo@teltier.com

  • Venue:
  • Journal of Network and Systems Management
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

The increasing complexity in size and heterogeneity of networks requires the development of network management tools that are less dependent on human intervention. Many tasks in network management consist of the coordination and execution of multiple activities across different network platforms. In the first part of this paper we introduce a language to write Distributed Action Plans (DAPs). A DAP is meant to specify one of these distributed network management tasks. To run these plans we use policy agents. These are agents that can be specified in \cal{PDL}, an event-driven programming language developed at Bell Labs to perform policy-based network management. The second half of the paper shows how DAPs are translated into \cal{PDL} and describes the policy agent system that executes these DAPs.