How to reduce server count by 50% and deliver better service


By Ian Lumb | August 21, 2014 |



Add a project, add a few servers. Add a product, add a few more servers. Add a team, add a gaggle of servers. Rinse and repeat on the scale of an industrial manufacturer with an international presence, and you have a proliferation of servers to the tune 13,000. Large numbers mask subtle, yet potentially serious concerns. For example, unmanaged servers present security vulnerabilities that can expose entire organizations and prove costly.

In a new case study, we introduce you to one of our customers who confronted the challenge of server proliferation. Their approach includes the creation of a small number of mega datacenters. They populated each datacenter with modern equipment and managed it with Bright Cluster Manager. Bright is a perfect fit on numerous levels. For example, Bright allows this industrial manufacturer to choose equipment from multiple vendors - equipment that includes the latest coprocessors and accelerators. Because Bright excels at provisoning, monitoring, managing and reporting via an easy-to-use GUI, well maintained servers are the new norm.

Here's the kicker: Server count is reduced by 50% while utilization triples to 90%. Yes, this is an actual example of doing more with less.

It gets even better. The company's engineers consistenly report improved service, as compared to when they managed servers themselves. No doubt they're also better able to focus on engineering, rather than IT Infrastructure. Because the mega datacenters aggregate large numbers of servers, parallel computing at scale is now a possibility. 

You don't need to be a multinational corporation to benefit from consolidation.

To learn more, please download our case study

describe the image


High Performance Computing eBook