Embryonic models for self-healing distributed services

Abstract

A major research challenge in distributed systems is the design of services that incorporate robustness to events such as network changes and node faults. In this paper we describe an approach - which we refer to as EmbryoWare - that is inspired by cellular development and di.erentiation processes. The approach uses "arti.cial stem cells" in the form of totipotent nodes that di.erentiate into the di.erent types needed to obtain the desired system-level behaviour. Each node has a genome that contains the full service speci.cation, as well as rules for the di.erentiation process. We describe the system architecture and present simulation results that assess the overall performance and fault tolerance properties of the system in a decentralized network monitoring scenario. © Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering 2010.