Self-stabilizable symbiosis for cloud data center applications: A game theoretic perspective

This paper describes and evaluates a self-stabilizable adaptation framework for data center applications. The design of the proposed architecture, SymbioticSphere, is inspired by key biological principles such as decentralization, natural selection, emergence and symbiosis. In SymbioticSphere, each data center application consists of application services and middleware platforms. Each service and platform is designed as a biological entity, analogous to an individual bee in a bee colony, and implements biological behaviors such as energy exchange, migration, replication and death. SymbioticSphere allows services and platforms to (1) adaptively invoke their behaviors according to dynamic network conditions and (2) autonomously seek stable behavior invocations as equilibria (or symbiosis) between them. A symbiosis between a service and a platform is sought as a Nash equilibrium in an extensive-form game. Simulation results demonstrate that SymbioticSphere allows services and platforms to successfully adapt to dynamic networks in a self-stabilizable manner.