Boost Productivity and Avoid Costs by Deploying Mixed Docker and HPC Workloads on Kubernetes
With the increased popularity of microservices architectures, organizations of all kinds are taking advantage of container management platforms like Kubernetes. By harnessing ready-to-run containers, users can evolve application services faster, share them more easily, and move seamlessly from test to production. But what happens when you are also a heavy HPC site supporting hundreds of users running analytic workloads?
This was exactly the challenge faced by the Institute for Health Metrics and Evaluation (IHME) an independent population health research center at the University of Washington. To support a growing number of Kubernetes applications as well as their compute intensive simulations running a 20,000 core Univa Grid Engine cluster, IHME turned to Navops Command, an innovative solution that extends the capabilities of Kubernetes allowing containerized and non-containerized workloads (including HPC workloads) to coexist seamlessly.
With Univa’s Navops Command, IHME can contain costs, improve utilization and pave the way to a modern container environment where they can enhance applications at their own pace.
IHME realized the following benefits with Navops Command:
Support for both containerized and non-containerized workloads
Improved cluster capacity and utilization
Improved ability to guarantee service levels
Minimal impact to existing HPC and container environments
Navops is a suite of products that enables enterprises to run containerized microservices applications and non-containerized workloads more efficiently only Kubernetes. With Navops Command, Grid Engine is deployed as a service, providing a seamless experience for HPC users while bringing advanced, enterprise-proven policy management to Kubernetes container orchestration. In short, Navops Command brings superpowers to your Kubernetes cluster and HPC workloads.