Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Evaluating Quality of Service Traffic Classes on the Megafly Network

Conference ·

An emerging trend in High Performance Computing (HPC) systems that use hierarchical topologies (such as dragonfly) is that the applications are increasingly exhibiting high run-to-run performance variability. This poses a significant challenge for application developers, job schedulers, and system maintainers. One approach to address the performance variability is to use newly proposed network topologies such as megafly (or dragonfly+) that offer increased path diversity compared to a traditional fully connected dragonfly. Yet another approach is to use quality of service (QoS) traffic classes that ensure bandwidth guarantees. In this work, we select HPC application workloads that have exhibited performance variability on current 2-D dragonfly systems. We evaluate the baseline performance expectations of these workloads on megafly and 1-D dragonfly network models with comparably similar network configurations. Our results show that the megafly network, despite using fewer virtual channels (VCs) for deadlock avoidance than a dragonfly, performs as well as a fully connected 1-D dragonfly network. We then exploit the fact that megafly networks require fewer VCs to incorporate QoS traffic classes. We use bandwidth capping and traffic differentiation techniques to introduce multiple traffic classes in megafly networks. In some cases, our results show that QoS can completely mitigate application performance variability while causing minimal slowdown to the background network traffic.

Research Organization:
Argonne National Laboratory (ANL)
Sponsoring Organization:
USDOE Office of Science; USDOE National Nuclear Security Administration (NNSA)
DOE Contract Number:
AC02-06CH11357
OSTI ID:
1574767
Country of Publication:
United States
Language:
English

Similar Records

An Evaluation of the Effect of Network Cost Optimization for Leadership Class Supercomputers
Conference · Fri Nov 01 00:00:00 EDT 2024 · OSTI ID:2538101

Union: An Automatic Workload Manager for Accelerating Network Simulation
Conference · Tue Dec 31 23:00:00 EST 2019 · OSTI ID:1828139

Performance trade-offs in reconfigurable networks for HPC
Journal Article · Wed May 11 00:00:00 EDT 2022 · Journal of Optical Communications and Networking · OSTI ID:1874993

Related Subjects