Are We Headed Toward Lights-Out Data Centers?

Aug. 6, 2021
Traditional networks tend not to be as scalable, agile or flexible as they need to be to support automation and lights-out data centers. Shad Secrist, Data Center Pre-Sales Solutions Engineer at Belden, discusses leaf-spine architecture in lights-out data centers. 

In this edition of Voices of the Industry, Shad Secrist, Data Center Pre-Sales Solutions Engineer at Belden, discusses leaf-spine architecture in lights-out data centers. 

Shad Secrist, Data Center Pre-Sales Solutions Engineer at Belden (source: Belden)

When someone mentioned a lights-out data center a decade ago, they were likely talking about a passive data center.

For redundancy and disaster recovery purposes, some organizations implemented an active-passive architecture: They built an active data center to serve as the primary location and a passive data center within a few hundred miles of that location. If the active site went down, then the passive data center was on standby to take over and avoid downtime.

Today, however, a lights-out data center (also called an “unmanned data center”) is just what its name implies: a fully automated facility that can operate in the dark without onsite staff. The lights can literally be turned off to save the operator money. Instead of engineers and operations managers being onsite, data center services and equipment are monitored and managed remotely.

This means that conditions don’t necessarily have to be suitable for long-term occupancy. Decisions about things like location, aisle layouts, temperature and server rack heights may not need to put human operators at the forefront—instead, the focus can be on maximizing square footage. When people are needed onsite (for hardware change, for example), the data center sends an alert to let them know.

Lights-out data centers are a good example of what we call hybrid edge data centers. They support a mixed approach to computing, storage, security and carrier services for efficient and cost-effective data management, storage and analysis. They deliver latency-sensitive data close to users and integrate with core or centralized applications in public clouds or corporate data centers. Instead of supporting only outward-bound data, they also support bidirectional data.

Lights-out data centers may not be mainstream yet (although many IT departments already work with technology like software-defined networking [SDN] and virtualization), but COVID-19 showed many operators what they might look like. In many cases, the pandemic proved that data centers could still operate with much less human involvement than originally thought.

Layer 1 Automation through Leaf-Spine Architecture

The data center’s traditional approach to networking has not kept up with technology. As applications and goals change, a more flexible topology is needed. Traditional networks tend not to be as scalable, agile or flexible as they need to be to support automation and lights-out data centers.

As data centers move toward automation, many get stuck as they attempt to automate layer 1 (the physical layer), which “plays” with cabling, connectivity and local area networks (LAN). Software-defined networking (SDN) is one way to facilitate automation at this level. While IT operations teams have been providing on-demand services and change requests, layer 1 networking has remained stubborn.

To facilitate automation, you can eliminate cross-connects from meet-me rooms and instead create a full mesh-network topology with leaf (aggregation) switches (which connect servers and storage) and spine switches with high port density (which connect leaf switches). This is called leaf-spine architecture.

The spine switches in the main distribution area don’t connect to one another. Instead of running multiple point-to-point connections, each leaf switch connects with each spine switch. The spine provides traffic-forwarding options for leaf switches. Traffic only moves from an ingress leaf switch to a spine switch and out to an egress leaf switch, which decreases bottlenecks and latency for faster data transmission.

With cloud infrastructure on the rise, east-west traffic (traffic that moves from server to server) is growing. This kind of traffic requires low latency to support time-sensitive, data-intensive applications. Leaf-spine architecture supports this effort by making sure traffic is always the same number of hops from its next destination.

As we anticipate deployment of leaf-spine architecture to support high-speed, low-latency communications, structured cabling will be at the heart of making it possible.

Shad Sechrist is Data Center Pre-Sales Solutions Engineer at Belden. He has more than 20 years of experience in the electrical and electronic manufacturing industry, including data center design, building infrastructure consulting and network design. Belden designs, manufactures and markets a comprehensive portfolio of networking, security and connectivity technologies and products across a variety of industrial, enterprise and professional broadcast markets. Learn more here about the Belden cabling and connectivity solutions that help you improve space utilization and support low latency in your data center.

About the Author

Voices of the Industry

Our Voice of the Industry feature showcases guest articles on thought leadership from sponsors of Data Center Frontier. For more information, see our Voices of the Industry description and guidelines.

Sponsored Recommendations

Optimizing AI Infrastructure: The Critical Role of Liquid Cooling

In this executive brief, we discuss the growing need for liquid cooling in data centers due to the increasing power demands of AI and high-performance computing. Discover how ...

AI-Driven Data Centers: Revolutionizing Decarbonization Strategies

AI hype has put data centers in the spotlight, sparking concerns over energy use—but they’re also key to a greener future. With renewable power and cutting-edge cooling, data ...

Bending the Energy Curve: Decoupling Digitalization Trends from Data Center Energy Growth

After a decade of stability, data center energy consumption is now set to surge—but can we change the trajectory? Discover how small efficiency gains could cut energy growth by...

AI Reference Designs to Enable Adoption: A Collaboration Between Schneider Electric and NVIDIA

Traditional data center power, cooling, and racks aren’t sufficient for GPU-based servers arranged in high-density AI clusters...

Courtesy of AMI
Image courtesy of AMI

The Next Era of AI Data Centers: Why Device-Level Management Matters

Rami Radi, Senior Product Manager and Solution Architect at AMI, explains why a non-intrusive, IT-centric and heterogeneous approach to data center infrastructure management is...

White Papers

Dcf Imdcwp Cover 2023 01 11 17 19 43

Infrastructure Planning Report - EMEA - Frankfurt

Jan. 11, 2023
In this white paper, Iron Mountain Data Centers provides an overview of the German colocation market. It explores strengths and weaknesses of the market as well as the latest ...