Coordinated Self-Adaptation in Large-Scale Peer-to-Peer Overlays

Self-adaptive systems typically rely on a closed control loop which detects when the current behavior deviates too much from the optimal one, determines new optimal values for system parameters, and applies changes to the system configuration. In decentralized systems, implementing each of these steps is challenging, especially when nodes need to coordinate their local configurations. In this paper, we propose a decentralized method to automatically tune global system parameters in a coordinated manner. We use gossip-based protocols to continuously monitor system properties and to disseminate parameter updates. We show that this method applied to a decentralized resource selection service allows the system to quickly adapt to changes in workload types and node properties, and only incurs a negligible communication overhead.