Joint Effects of Application Communication Pattern, Job Placement and Network Routing on Fat-Tree Systems

Abstract

Among the high-radix and low-diameter networks, fat-tree topology is commonly used in high-performance computing (HPC) and datacenter systems. Resource and job management on HPC systems is critically important to mitigate application interference in order to achieve high system performance and utilization. Preliminary studies have shown the effect of job placement on parallel scientific applications performance in fat-tree network. In this work we explore the joint effects of job placement and network routing aware of applications communication pattern on fat-tree system. Applications can be classified into various groups according to the communication patterns. We further combine various job placement policies and routing algorithms and create six different configurations. The system performance is analyzed using communication, hops, traffic, and saturation data by performing fine-grained high-fidelity discrete event-driven simulation. Initial experimentation shows that the performance of HPC applications not only is related with the communication pattern, but also relies on the job placement and network routing on fat-tree systems.