Efficacy Analysis of a SDN-enhanced Resource Management System through NAS Parallel Benchmarks

Abstract

In the field of social science, a variety of high-performance computing simulations such as the Monte Carlo simulation and the Multi-agent simulation must be efficiently performed to deal with social scientific big data. To facilitate Rev Socionetwork Strat (2014) 8:69-84 70 social scientists in performing their own analysis against such big data, the information infrastructure for social science must be equipped with a core technology that efficiently and effectively leverages limited resources available on the information infrastructure. From such a perspective, a new type of job management technology, which treats not only computational resources such as the Central Processing Unit (CPU) and memory, but also network resources unlike traditional job management, is investigated in this paper. A cluster system with a fat-tree topology interconnect is conventional cluster architecture these days. For this investigation, the National Aeronautics Space Administration Advanced Supercomputing, USA (NAS) Parallel Benchmarks, which contain computation patterns often observed in social scientific simulations, are used to assess the efficacy of the resource allocation by our proposed job management technology on a cluster system with a fat-tree topology interconnect.

Dynamic Traffic Control of Staging Traffic on the Interconnect of the HPC Cluster System

“…Based on this concept, an SDN-enhanced job management system (JMS) [22], [23], which is a job scheduler to manage not only the computing nodes but also the interconnect to improve the job throughput, and SDN-enhanced MPI [24], [25], which enhances MPI protocol using SDN to improve the performance of inter-node communication, have been proposed. In SDN-enhanced JMS [22], [23], the job scheduler allocates computing nodes to a job based on the topology of the interconnect so that as many of the computing nodes as possible are connected in one hop, and then allocates the path with the most available bandwidth to a pair of each computing node allocated to the job for improving the inter-node communication performance. In SDN-enhanced MPI_Bcast [24], MPI_Bcast, which is one of the MPI functions, has been enhanced using SDN so that the amount of the traffic caused by MPI_Bcast is reduced.…”

Softening Up the Network for Scientific Applications

“…In general, the performance of parallel and distributed computation is greatly affected by communication due to network congestion points or bottlenecks [24]. To tackle these bottlenecks, a multiple-path network usually provides more than one path between any pair of computing nodes.…”