Easily Manage a Diverse Set of Workloads with Bright Cluster Manager


By David Dean | March 12, 2019 | Bright Cluster Manager, Containers, VM



High-Performance Computing has undoubtedly changed in the last few years and administrators are frequently tasked with supporting unusual request. Gone are the days of all HPC cluster users being script savvy Computer Scientists who speak BASH as fluently as they do English.

As High-Performance Computing has matured, so too has the diversity of its use. Clusters are still used for finite element analysis and Monte Carlo simulations, but today they are being used more and more by Statisticians, Psychologists, Sales and Marketing Teams and a variety of “Data Scientists” who analyze large amounts of data and create neural networks for everything  from predicting the future to driving our cars. This expansion in use has created a challenge for today’s cluster administrator. They now have requests to support a variety of applications; some are homegrown and built on a laptop, others are commercially developed and may only run in Ubuntu (when the IT department has standardized on RHEL).  Some need to run on bare metal for speed, others in the cloud, and a few in containers and VMs.

Bright Cluster Manager can handle all of these scenarios. Bare metal jobs, VMs in OpenStack, AWS or Azure, Docker, and Singularity containers, all can run on a Bright managed cluster.  With Bright Cluster Manager, a wide variety of images with countless applications can easily be provisioned onto bare metal. In minutes, OS images stored in your library can be moved from one bare metal system to another or to VMs that are on-premise or in the cloud. If your Data Scientist has cooked up his own recipe and baked it in his own special container, no problem.  Bright can manage hundreds of those containers with Kubernetes or just one container that gets added to the workload manager queue. If your cluster isn’t big enough to handle that time-critical workload that has executive focus, seamlessly extend your cluster into AWS or Azure adding virtual resources to get the job done. 

If you’re interested in learning more about how Bright can make it easier to manage a diverse set of workloads, we’d love to speak with you.