Generative Policy model for Autonomic Management

Abstract Policy based Management has been used for autonomic management in several systems, but current approaches have used a model of policies that follow the event-condition-action paradigm or variants thereof. While proven successful in many contexts, these systems nevertheless require a manager to predict future conditions and define the applicable rules in advance. A higher level of autonomic behavior can be enabled if the managed system could be allowed more flexibility in its actions, while maintaining the requirements for consistency and compliance that have led to the successes of current policy based management paradigm. In this paper, we present a new approach for policies - which enables managed systems to take more autonomic decisions regarding their operations.
  • Dinesh Verma (IBM US)
  • Seraphin Calo (IBM US)
  • Supriyo Chakraborty (IBM US)
  • Elisa Bertino (Purdue)
  • Chris Williams (Dstl)
  • Jeremy Tucker (Dstl)
  • Brian Rivera (ARL)
Date Aug-2017
Venue International Workshop on Distributed Analytics Infrastructure and algorithms for multi organization federations