Decentralized Probabilistic Auto-Scaling for Heterogeneous Systems

Scalability has become a de facto non-functional requirement for today’s Internet applications that start small and aim to become a huge success. The ability of the system to automatically scale is required by the dynamic nature of the workload experienced by these applications. In this context, the DEPAS (Decentralized Probabilistic Auto-Scaling) algorithm assumes an overlay network of computing nodes where each node probabilistically decides to shut down, allocate one or more other nodes, or do nothing. DEPAS was formulated, tested, and theoretically analyzed for the simplified case of homogenous systems. In this paper, we extend DEPAS to heterogeneous systems. Thus, we provide a new formula for computing the node-level addition probability and evaluate it both theoretically and experimentally. Keywords-auto-scaling; decentralized computing; randomized algorithms; cloud computing