Maintenance of Monitoring Systems Throughout Self-healing Mechanisms

Monitoring is essential in modem network management. However, current monitoring systems are unable to recover their internal faulty entities forcing the network administrator to manually fix the occasionally broken monitoring solution. In this paper we address this issue by introducing a self-healing monitoring solution. This solution is described considering a scenario of a monitoring system for a Network Access Control (NAC) installation. The proposed solution combines the availability provided by P2P-based overlays with self-healing abilities. This paper also describes a set of experimental evaluations whose results present the tradeoff between the time required to recover the monitoring infrastructure when failures occur, and the associated bandwidth consumed in this process. Based on the experiments we show that it is possible to improve availability and robustness with minimum human intervention.