Autonomic Management of Cloud Neighborhoods through Pulse Monitoring

Abstract

This paper reports on autonomic computing research, including the development of a self-* proof of concept, for a cloud based environment. It monitors administrative boundaries from within an autonomic manager, with each manager operating in a peer-to-peer mode and utilizing a pulse monitor. The prototype was developed in Java utilizing SNMP to demonstrate the manager's self-situation and environment-awareness of the current state of the whole neighborhood and proves the feasibility of communicating the health of the neighborhood to peer managers using an XML pulse concept. Each manager houses the functionality to enact changes to their neighborhood using SNMP based rules. This enables the capability to provide self-healing, self-configuring, self-optimizing and self-protection to network neighborhoods within cloud computing.