Hands-on Report: Open Networking approach to modern hyperscale data center

Ed. Note: This post was authored by DJ Spry, Principal Engineer, Dell Open Networking

In the three short months since I joined Dell, there are two questions I frequently hear from customers about Open Networking (ON):

  • Why does it matter?
  • How can we leverage it?

Let’s start with “Why does it matter?”

Dell is providing a disaggregated hardware and software ecosystem where customers have the choice of best-of-breed Ethernet switches, innovative third-party operating systems, and open standards-based applications and tools.  Companies adopting this paradigm shift will be able to realize lower costs, increased agility, simpler scalability and a network architecture that is far more flexible. 

The “How can we leverage it?” question requires more detailed discussion as to specific use cases. For example:

  • Are you leveraging any Cloud Management Systems?
  • OpenStack, VMware, or both?
  • How Linux savvy is your staff?
  • Network Virtualization and Overlays?

Those are just a few clarifying questions to determine what solution would best fit their business needs.   However, more times than I can count the conversation circles around how their legacy network is too complex to manage and simply not able to handle the workloads of highly virtualized environments.

Recently the Enterprise Strategy Group (ESG), a leading technology research group, spent some time evaluating and testing Big Switch Networks Big Cloud Fabric software running on Dell Open Networking switches. 

This report does an outstanding job of not only explaining the basic operation of this NOS but also specific use cases around analytics, cloud management system integration, application-centric provisioning, multi-tenant provisioning, and programmatic capabilities. 

Here are a few examples pulled from the report that highlight they simplicity and flexibility of the system:

  • ESG Lab verified that Big Cloud Fabric required just a few keystrokes and took less than 15 minutes to go from bare hardware to a fully functioning enterprise data center network fabric.
  • The simplicity of the BCF controller GUI was demonstrated by the automatic laying out and routing of the tenant and segment mapping.
  • These graphics enabled ESG Lab to simply and quickly gain understanding of the logical configuration of the fabric and the tenants without having to read and interpret command line configuration statements, or navigate through each configuration page in the GUI

This report highlights how combining just one of our NOS’s can provide savings in both CapEx and OpEx by delivering a simple, scalable, and elastic network.  To read more you can check out the report here.

About the Author: Adnan Bhutta