Epidemic-Style Global Load Monitoring in Large-Scale Overlay Networks

Peer-to-peer systems are highly dynamic systems, with permanent changes in their configurations, as peers may join and leave the system with no restriction or control. This makes monitoring an important element for several applications, especially for fault management: fault detection and fault recovery. As the system may have a large number of nodes, we need a scalable algorithm, able to guarantee a fast convergence no matter what the structure of the network is. Epidemic-style or gossip-based algorithms offer solutions for various topics in large-scale distributed systems. Even though they present many advantages due to their property of continuously spreading the information across the system in a reactive and proactive fashion, several problems such as the total number of messages exchanged between peers and the number of rounds required for ensuring the convergence of the algorithm appear. In this paper we present a gossip-based algorithm for monitoring large-scale distributed systems and we analyze its efficiency in a simulated environment provided by OverSim.