The Network Doesn’t Build Itself: Inside SCinet Staging

Before thousands of attendees and exhibitors arrive at SC26 and researchers begin pushing the boundaries of high-performance networking, another kind of work is well underway.

Inside a staging area in Chicago, SCinet volunteers are busy installing equipment, connecting fiber, configuring switches, deploying services, testing systems, and troubleshooting the countless details required to bring one of the world’s most ambitious temporary networks to life. Behind this effort are nearly 18 specialized teams, each contributing a critical piece of the infrastructure that will power SC26. From physical networking and security to automation, systems administration, and beyond, their work is closely interconnected. What happens during staging sets the foundation for everything that follows.

To find out what it takes to build a network of this scale and why staging is so important, we caught up with three of the teams working behind the scenes, DevOps, WAN, and Routing. Their stories offer a glimpse into the technical challenges, coordination, and collaboration happening across the larger SCinet operation. While each team has its own responsibilities, their work illustrates something fundamental about staging: building SCinet is not a series of isolated tasks. Instead, it is a coordinated effort, where every connection, configuration, and system depends on the work of others.

DevOps: Before the First Cable Is Connected

A walk through the SCinet staging area is marked by racks of equipment, bundles of fiber, and engineers working side by side. Yet much of the work that makes those connections possible begins long before anyone plugs in a cable. This is where the DevOps (Development and Operations) team comes in.

Led by Chairs Karen Lopez (National Laboratory of the Rockies (NLR), Ross Lindsay (Georgia Institute of Technology), and Doug Southworth (Texas Advanced Computing Center), DevOps serves as the system administration backbone for SCinet, managing essential services, including DNS, DHCP, virtual machines, VPNs, network measurement tools. The team manages the “Roadie Rack,” which hosts essential infrastructure services, and deploys virtual machines for other teams. One major DevOps accomplishment this year was modifications to the Connection Request System, the interface that exhibitors use to request fiber connections to the SCinet network.

Nautobot, SCinet’s “Network Source of Truth,” is where the network’s planned infrastructure and connections are documented and tracked, making it the centerpiece of DevOps’work. Another major accomplishment for SC26 was improvements to VM provisioning using Proxmox, the Nautobot patching application. During staging, DevOps turns that information into tools that the other teams can use. The Nautobot patching application helps engineers translate planned connections into physical cabling, while automation jobs make it possible to process large numbers of changes efficiently.

“Before a single cable is plugged in, the entire network is planned, documented, and tracked in one place,” Lopez, a network automation engineer with the Department of Energy’s NLR, explains.

The challenge? Almost everyone needs something from DevOps. Lopez’s team addresses that pressure by automating work wherever possible and prioritizing the requests that other teams depend on. It is a reminder that staging is not merely about installing hardware. It is building the systems that allow hundreds of individual tasks to move forward as one coordinated effort.

WAN: From Network Plans to Physical Connections

Once the network is mapped out, someone has to make those connections real. For the WAN (Wide Area Network) team, led by Henry Willoughby (North Point Ministries), Chris Skaar (University of Illinois), and Camille Alfs (Internet2), that means turning diagrams, connection records, and months of coordination into physical infrastructure. The team’s responsibility extends well beyond the conference center. WAN coordinates circuits connecting SCinet to research and education networks and institutions across the country and, in some cases, around the world.

During staging, the WAN team members are working from two fronts. Outside the venue, circuits and cross-connects are being established at points of presence, where fiber connections link SCinet to external networks. Inside the staging area, engineers are installing equipment and patching connections between network racks, patch panels, and the infrastructure that carries those external circuits into the building. It is meticulous, hands-on work. Every connection must be in the right place, and each circuit needs to be tested.

From carefully routing fiber cables to troubleshooting connections together, WAN engineers work through the details that make a reliable network possible. Willoughby, a network engineer with North Point Ministries, describes one of the more exciting moments of staging – when a distant connection finally comes to life.

“There’ll be a point on Thursday where it’s like, ‘Oh, wow, we see light from Los Angeles,’” he says. “‘Oh, wow, we see light from Washington, Virginia area.’”

Those moments represent far more than a successful fiber connection. They are the result of months of coordination among network engineers, carriers, contributors, and institutions. The WAN team also works closely with Routing, DevOps, and Network Security, ensuring the physical connections support everything from research demonstrations to the conference’s everyday Internet connectivity.

This year’s location for the SC Conference adds another dimension to the challenge. Willoughby says Chicago’s central position among research and education institutions has contributed to particularly strong interest and connectivity requests. From his perspective, the SC26 network is larger than ever. That makes staging even more important. With so many circuits and systems coming together, there is little room for guesswork.

As Willoughby puts it, the team is responsible for “getting the network wired up and basically connecting the world’s fastest temporary network to places around the country and, in some cases, around the world.”

Routing: Making the Network Come Alive

Physical connections are only part of the equation. Once equipment is connected, it must be configured, managed, and able to communicate across the network. This is where the Routing Team takes the lead.

According to Annalicia Duran, Routing Team Co-Chair from Internet2, one of the team’s most critical staging priorities is bringing up the out-of-band management network. This network allows engineers to access, configure, and troubleshoot devices remotely without having to physically connect to each piece of equipment. Such capability is especially important because there is a gap between staging and the SC Conference’s setup and show weeks. The infrastructure must remain accessible even when the teams are not physically together.

Engineer connecting fiber cables while team members inspect patch panels.

During staging, Routing is installing top-of-rack management switches, establishing connectivity, and coordinating external peering with commodity Internet and research and education network providers.

The team has already brought up one commodity Internet peering link and established management access to top-of-rack management switches. Before the staging ends, the team expects the out-of-band management network to be fully functional for all devices and SCinet teams. Getting there requires close collaboration with DevOps and Automation to ensure device information in Nautobot is accurate and can be used for automated configurations. Routing also works directly with WAN to make sure physical connections align with the planned network topology.

And, as with any major technical build, unexpected challenges are part of the process.

This year, one challenge has been securing the right quantity and types of optical transceivers needed for the network. The team is addressing shortages by working with vendors, reviewing existing inventory, and borrowing equipment from partner teams. It is the kind of problem-solving that rarely makes it into the conference spotlight but can make a significant difference in keeping a complex deployment on schedule.

Behind this effort is also a year of preparation. Duran credits fellow Routing leads Keith Vaughan (NLR), Jamie Bashore (Florida State University), and Kim Schjang (A.O. Smith) for their work designing the network, documenting workflows, and training team members ahead of staging.

Three Teams, One Big Effort

The work of DevOps, WAN, and Routing offers just one window into the much larger staging operation.

Across nearly 18 teams, volunteers are tackling different pieces of the same complex challenge. Some are building physical infrastructure, others are preparing services, securing systems, automating configurations, or ensuring everything will operate as expected when SC26 begins.

What connects them all is a shared goal of transforming months of planning into a reliable, high-performance network in an extraordinarily short period of time.

This is the crux of what makes staging so essential. Beyond a week of assembling equipment, staging is where plans are put into action, dependencies are worked through, problems are solved, and teams come together to ensure the network is ready before the first attendee walks through the doors at McCormick Place.

This is just a glimpse of what happens during staging. By showtime, much of that effort will be invisible. However, behind every successful connection will be the collective work of dedicated SCinet volunteers across all of these teams, whose SCinet stories are worth telling.