Skip to content

HSRN Kubernetes Cluster

Introduction

This cluster is run by the High-Speed Research Network group (HSRN), part of High Performance Computing at New York University. It is available for teaching and research, currently in a pilot phase. It offers access to Kubernetes, an open-source container orchestration system that you can use to run experiments or applications on the cluster as containers.

Use of this system relies on you understanding Kubernetes and how to use it. There is extensive introductory documentation that can help you get started if you are new to it. Note that Kubernetes can be harder to use than traditional HPC such as SLURM. You will have to provide detailed configuration for your desired workloads, in YAML format. In particular the filesystem of a container is not persistent by default, if you don't explicitly mount a persistent volume you might lose data! So do read the documentation and try a few examples on the cluster if you are just starting out with Kubernetes.

Where to get support

Please email hsrn-support@nyu.edu with questions or issues.

Hardware overview

  • Dell PowerEdge machines, 36 total
  • Distributed in 7 different NYU buildings
  • 724 cores
  • 11027 GB of RAM
  • 14 NVIDIA GPUs

See the status page for current utilization.