How Federal Agencies Use Virtualization To Prevent Downtime

Virtualization platforms proactively monitor server health, automatically balancing workloads across servers to prevent bottlenecks. If a server degrades or fails, VMs live-migrate to healthy servers to prevent downtime. Administrators can also perform upgrades and maintenance without taking systems offline. And if a data center goes down entirely, virtualization platforms can replicate VMs and data and failover to a secondary data center or the public cloud for disaster recovery.

“Virtualization is still the data center control plane,” says analyst Matt Kimball of Moor Insights & Strategy. “It delivers resilience, which is absolutely critical for any military or continuous operation kind of environment.”

Ensuring Uptime With HCI

The Navy’s Nutanix HCI server appliances are spread across several server rooms in an active-active configuration with redundant power and fiber-optic cabling for failover. Running on a hypervisor, the systems sustain a diverse array of medical capabilities necessary to deliver world-class patient care, Belcher says.

Virtualization helps prevent downtime. If one cluster goes offline, workloads automatically shift to another cluster elsewhere on the ship, so doctors and nurses never lose access to medical systems.

“We have primary and failover clusters that support the rapid pickup of services in the event we lose one side,” Belcher says. The transition is seamless, so users experience no disruption.

LEARN MORE: Virtualization is the foundation of modern hybrid infrastructure.

Data is also automatically mirrored, protecting patient records from hardware failures. Live migration allows the IT staff to perform maintenance and patching without impacting patient care. Ease of management is essential for maximizing the operational efficiency of the ship’s IT team. 

“The administrative overhead of managing the system is low, while resiliency and capability are high,” he says. “This allows our teams to focus on issues and problems as they arise. We’re not locked into constant monitoring or tuning of systems.”

Modernizing With Virtualization

The Federal Aviation Administration (FAA) is modernizing its air traffic control infrastructure with the Terminal Flight Data Manager (TFDM), a real-time system that replaces paper flight strips and shares electronic flight data among controllers, airlines and airports simultaneously, which helps controllers make faster, more informed decisions and improves air safety and air traffic flow.

Downtime is not an option in air traffic control operations — and virtualization is the technology foundation that helps ensure uptime.

The FAA implemented TFDM at its 17th site, Chicago Midway International Airport, in mid-April, and plans to deploy it to 89 towers by 2028. Each TFDM deployment runs on an on-premises data center scaled to site-specific needs, an FAA official says.

The infrastructure is powered by a mixed server environment — up to five physical servers running VMware vSphere and hosting up to 10 VMs, alongside up to 10 bare-metal Red Hat Enterprise Linux servers. The FAA also standardized on NetApp and Hewlett Packard Enterprise storage hardware and HPE Aruba switches for networking.

Automated failover and live migration keep TFDM running even if hardware fails, critical for a system where controllers are managing air traffic. Virtualization simplifies disaster recovery, as entire system environments can be replicated and restored across multiple data centers far more quickly than with physical hardware alone, the FAA says.

Beyond reliability, virtualization also supports latency-sensitive operations through intelligent load balancing, which dynamically allocates computing power across servers to prevent performance bottlenecks.

The FAA also leverages dedicated VMs to keep TFDM running smoothly. A Maintenance Control Workstation VM allows tech ops staff to monitor system health and perform maintenance, while a separate VM allows technicians to test and validate system changes before deploying them into production.

“This creates a flexible, highly available environment where maintenance and upgrades can occur without interrupting the critical flow of flight data,” the FAA official says.

Strengthening Disaster Recovery

Except for legacy mainframe systems still in use, virtualization has become essentially universal across federal agencies, says Daniel Chenok, executive director of the IBM Center for The Business of Government.

Adoption of virtualization has evolved from basic server consolidation and enabling more efficient use of hardware to modernizing operations and supporting the kind of flexible, resilient infrastructure required for emerging technologies such as AI.

“In a technologically advanced world, with AI here and quantum computing on the horizon, the infrastructure for the applications that enable those tools to succeed needs to be flexible,” Chenok says. “Virtualization is a critical enabler of the government being able to adopt modern technology.”

When Tom Soderstrom became NASA Jet Propulsion Laboratory’s first IT CTO in 2006, leadership asked him to spend less on IT infrastructure and more on engineering insights and the future. That prompted him to experiment with both virtualization and the cloud.

Soderstrom used virtualization in the cloud for compute-intense science workloads and to bolster disaster recovery. JPL initially used the cloud to back up on-premises systems. Then, the lab made the cloud primary, with on-premises as backup. Eventually, JPL ran primary workloads in one cloud environment with another cloud instance as backup.

“This gave us full disaster recovery, and we could focus on-premises resources on workloads that could only run there and learn how to use more advanced ways of working in the cloud,” says Soderstrom, who left JPL in 2020 and is now at Amazon Web Services as an executive in residence.

Margie Graves, a senior fellow at the IBM Center for The Business of Government, saw similar benefits during her tenure as deputy CIO at the Department of Homeland Security, from 2008 to 2016. The technology strengthened resiliency for 24/7 operations across DHS, where downtime was not an option, she says.

“If anything did go bump in the middle of the night, you could fail over and maintain your uptime,” Graves says. “When hurricanes, floods and fires threatened infrastructure capability, having a virtualized environment to fail over was definitely some help in maintaining constant operations.”