Reimagining Data Center Networking with Cisco Nexus Hyperfabric

Back to basics in the data center

Two years ago, my team and I (later known as the Cisco Nexus Hyperfabric team) got the green light to rethink—from the ground up—where data center infrastructure solutions need to be. And then we built it. Fast.

But first, we rallied scores of customers, partners, and experts to gain broader insights. We’re forever grateful for the time and guidance they provided during deep-dive sessions.

One thing stood out to me: from the first hour of the first meeting with the first advisor, it was all about the basics. At that first meeting, the topic was something their team deals with all the time: cabling and rack & stack issues. Then we got to the complexities of the network and all the protocols that made a fabric work.

Another person delivered this gem: “The CLI I learned 25 years ago I still use every day today—why?”

Our advisors really were clear on how we should do better. And we listened. Our team built Cisco Nexus Hyperfabric with these guiding principles in mind:

• Keep it simple
• Deliver a reliable outcome
• Own the process from design to day 2
• Data centers must be agile, not monolithic

Building a cloud-managed networking solution

We spent the time listening so that we could make it easier for you to build data center fabrics. At its core, Cisco Nexus Hyperfabric is practically a “fabric as a service” through automation and cloud management. You can use its cloud controller to easily design, deploy, and manage any number of fabrics located anywhere, spanning primary data centers, colocation facilities, and distributed data center edge sites. And this vertical stack solution comes with everything you need, including purpose-built hardware, software, cloud management, day-2 automation, and Cisco support.

Built for primary data centers of up to 1000 servers, Cisco Nexus Hyperfabric is also a great solution for distributed or edge data centers with 10 to 100 servers. It’s perfect for manufacturing or energy sites, hospitals, transport hubs, and the like.

All the benefits, none of the hassle

Based on input from our trusted partners and customers, we knew we needed to address the complex lifecycle of the data center. With Cisco Nexus Hyperfabric, we reinvent the IT operations lifecycle of the data center by simplifying each step of the process. You’ll get repeatable and predictable outcomes no matter who is using it: IT-generalist, application, or DevOps teams.

From build to delivery and on through implementation and management, the Hyperfabric lifecycle is truly unique and provides a seamless and simple process.

Imagine being able to design your data center fabric without spending a dime upfront. With Cisco Nexus Hyperfabric, anyone with a Cisco Connection Online (CCO) account can do just that. It’s like having a sandbox where you can play around with different configurations before committing to a purchase. These designs are more than just sketches; they’re comprehensive blueprints. They cover everything from the model and capacity of switches to cabling, airflow direction, Layer 2 and Layer 3 host port services, and even upstream routing connections.

And here’s the kicker: there are no unsupported blueprints. You can’t build something that isn’t supported, ensuring that your designs are always viable and ready for deployment. With Cisco Nexus Hyperfabric, managing your data center fabric has never been easier or more intuitive.

Once you’re ready to deploy, the switches are hardwired to call-to-cloud, meaning they automatically connect to the cloud. Once you claim and bind a switch to a fabric, it is automatically configured. It’s as if the switches have a mind of their own, seamlessly integrating into your network.

Innovative mesh deployments, effortless management

Cisco Nexus Hyperfabric supports traditional spine-leaf topologies, but it also offers something unique: mesh deployments, ideal for distributed and remote compute clusters. With just two switches, you can create a mesh fabric with 120 host-facing ports, providing flexibility and efficiency that stands out in the market.

One of the best parts? You don’t have to worry about the fabric’s internal network. All the complex underlay details like VXLAN, EVPN, BGP, VTEPs, and IPv4/6 addressing are hidden from view. You might say: “It just works.” And it allows you to focus on what matters without getting bogged down in the technicalities.

Forget about command-line interfaces or config files. Everything is managed in the cloud through a browser or, even better, via an API. This modern approach simplifies operations and keeps everything streamlined.

Monitoring and upgrades the easy way

For day-to-day operations, all the telemetry information you need is automatically collected. The system uses a new model called assertions to detect any issues when what’s deployed doesn’t match the blueprint. It’s like having a vigilant watchdog ensuring everything runs smoothly.

Upgrading your fabric is a breeze. Just pick a time, choose the switches that should be upgraded, and let the system handle the rest. If you need to adjust the fabric size, simply tweak your blueprint. The complexity of reworking the interior underlay is managed through automation, so you don’t have to worry about it.

Go on, give it a try

During Cisco Live Amsterdam, you can preorder Cisco Nexus Hyperfabric, with delivery later in the month. Why not give it a try? Until then, you can get more information on our website. Got more questions? We’ve got answers.

Share: