By Robert Stober | March 18, 2021 |
There is a lot of excitement around Bright Cluster Manager 9.1 providing an integration with vSphere, as it allows high-performance clusters to be created and managed without the need for skills and knowledge of vSphere.
VMware (vSphere) is the industry-leading commercial virtualization software used by organizations to increase the utilization of their computing infrastructure, allowing them to reduce costs by purchasing fewer servers and improve service to their staff by making it possible to quickly deploy virtual servers to teams when needed.
Many organizations running HPC, machine learning, and data analytics applications have an insatiable appetite for more and more computing capacity, which is throttled only by budgetary constraints.
The computing infrastructure that is virtualized with vSphere in most organizations is typically not utilized 24x7x365 the way that high-performance infrastructure is. As a result, there is often unused capacity in the servers running vSphere that could be used to provide virtualized high-performance clusters for HPC, machine learning, and data analytics applications when available.
Building high-performance clusters - whether physical or virtual in vSphere - is complex and difficult, and many cluster administrators are not skilled in vSphere.
Bright has had a cluster-as-a-service (CaaS) offering for many years, which allows you to spin up Bright clusters in AWS, Azure, and OpenStack within minutes. Bright Cluster Manager 9.1 saw the introduction of an integration with VSphere, which enables high-performance clusters to be created and managed without the need for skills and knowledge of vSphere.
This integration enables organizations to fully utilize their vSphere infrastructure (i.e., 24x7x365), saving money, quickly giving teams the ability to spin-up Clusters-as-a-Service, and accelerating innovation for the business by increasing the capacity of high-performance infrastructure that drives innovation.
Tapping into available vSphere capacity for high-performance needs will also reduce the need for organizations to use public cloud infrastructure, further reducing their costs and also eliminates the need to move data into the public cloud.