As part of its new research — 2018 Next-Gen Data Center Networking Report — the SDxCentral research team found that as data center operators move toward virtualization and other new technologies, cost is not the driving factor. Rather, manageability and agility come out ahead.
The new report combines data about software-defined networking (SDN) controllers with research about next-gen data center networking. “The data center is changing rapidly, and the concept of open hardware and software is also changing the landscape dramatically,” said Roy Chua, chief product officer and co-founder of SDxCentral. “The community has been pushing us to provide a comprehensive review.”
The primary business reasons that organizations are adopting new technologies are better manageability and velocity, according to an informal poll of about 60 members of the SDxCentral community. Twenty-seven percent said the “improved capabilities to configure, troubleshoot, and measure data center networks” was the biggest business driver for using next-generation technologies. Meanwhile 20 percent said it was the ability to “accelerate time to market.” Only 10 percent indicated that “reducing costs” and “feature richness” were the main business reasons their organization deployed new technologies.
When probed on which next-generation technologies they currently used in their data centers, 65 percent of respondents cited network virtualization solutions. Fifty-eight percent said they used vSwitch on hypervisors, and 56 percent had deployed SDN technologies and SDN controllers. And 26 percent were using a 2- or 3-stage CLOS Leaf-Spine architecture.
The research identified and delved into several trends in data center infrastructure, including:
- Design patterns that improve scalability and performance such as n-tier CLOS fabrics;
- The pervasive use of virtualization throughout the stack such as virtual machines (VMs), containers, and network functions virtualization (NFV).
- The use of SDN and programmable infrastructure.
- Aggressive data collection to feed infrastructure analytics.
- The creation of hybrid cloud infrastructure.
- Adoption of open hardware and software based on interchangeable commodity equipment.