
When a single router anomaly can cascade into a nationwide ground stop, the story is not just about a contractor on a ladder—it is about how modernization work interacts with legacy architecture in the most safety-critical communications network in the country.
The Short Version
- The outage began during an L3Harris copper-to-fiber maintenance job at a remote FAA site; the timing was direct and consequential.
- A router associated with the work spewed abnormal connection attempts—on the order of more than a thousand every few seconds—destabilizing communications and some radar feeds.
- The FAA restored operations by isolating the work and taking the offending functionality offline.
- The operational footprint—nationwide ground stop and thousands of disrupted flights—matches a broad FAA network instability, not an airline-side problem.
- L3Harris publicly said it is supporting the FAA investigation; ultimate fault between installer error and equipment failure had not yet been assigned.
What happened and why it mattered operationally
During a routine modernization action—replacing an aging copper telecommunications circuit with fiber at a remote FAA facility—L3Harris technicians were working in-line on the Federal Aviation Administration’s legacy telephony-and-data backbone when a router involved in the cutover began emitting abnormal traffic. Transportation Secretary Sean Duffy described the behavior as bursts exceeding 1,000 connection attempts every four to five seconds, sufficient to degrade controller-to-aircraft voice paths and destabilize some radar feeds across multiple facilities. That pattern is the textbook signature of a control-plane storm: a device repeatedly initiating sessions or advertising routes in a loop, overwhelming peer devices not built—or not segmented—to quarantine the flood.
With unstable communications in the National Airspace System, the safety calculus is straightforward. The FAA initiated a nationwide ground stop, then restored service after isolating the maintenance work associated with the malfunctioning equipment and disabling the behavior that triggered the storm. Duffy and FAA officials said there was no evidence of cyber or AI involvement—this was an equipment or installation failure surfacing in a fragile spot of the network.
The technical mechanism: how a localized maintenance slip propagated system-wide
At issue is the transition from legacy copper time-division multiplexing (TDM) circuits to Internet Protocol (IP) over fiber—an overdue, industry-standard move. In such cutovers, a router or media gateway sits at the seam, translating signaling and routing traffic into the IP domain. If that seam device is misconfigured—say, an errant keepalive timer, route advertisement loop, or authentication failure—it can hammer adjacent routers with repeated session attempts. If the device itself is faulty, similar symptoms arise: repeated retries, malformed packets, or rapid interface flaps. Duffy’s description—“more than 1,000 connection attempts every four to five seconds”—maps to an aggressive retry loop or broadcast storm that saturates control-plane processing budgets, which can then disrupt both IP voice and surveillance data distribution.
Why did this ripple beyond a single site? Two structural realities make the FAA’s environment vulnerable during live-line work. First, the legacy communications fabric is both central to voice and data distribution and long-lived, which means elements share interdependencies that predate modern segmentation standards. Second, voice switching and radar distribution have redundancy within facilities, but a cascade at the network layer can still impair multiple regions before isolation and rate-limiting kick in. The behavior here—facility instability first, then a broad stop for safety while teams cordoned the storm—fits that profile.
Responsibility: what is established and what remains under agency determination
The association with L3Harris’s maintenance is clear and uncontested in the published accounts: the network anomaly began during the company’s copper-to-fiber replacement at a remote FAA site. Duffy said the router at issue generated the abnormal traffic and that service recovered after isolating the work. L3Harris issued a statement the following day committing to support the FAA’s investigation into “the conditions that led to the Nov. 13 ground stop”.
Two possibilities remained under formal FAA determination at the time officials briefed the public: an installation or procedural error by technicians, or a defective router. Those are materially different failure modes—one human or process, the other hardware or firmware—but they are not mutually exclusive in practice, especially if a latent device defect is exposed by a corner-case configuration. What the public record does not yet provide are the device logs, NetFlow traces, or change-control artifacts that would disambiguate the initiating fault; nevertheless, the causal chain—maintenance session, router storm, isolation, recovery—holds as the operational narrative.
Scale of disruption: what the footprint tells us about the network
The event forced a nationwide ground stop lasting hours and triggered a second-day wave of cancellations and delays numbering in the thousands; by any operational measure, this was not a single-carrier problem but a shared network instability radiating across facilities. In aviation communications, scope is a diagnostic clue. Airline IT outages strand passengers; backbone anomalies deprive controllers of stable, high-confidence voice and data paths. The FAA’s decision to halt departures while controllers managed inflight traffic to safe landings reflects that difference and adheres to the system’s first principle: if communication integrity is in doubt, slow the system to safety.
How we got here: modernization colliding with legacy constraints
The FAA has spent years straddling two eras: sustaining a legacy telecommunication backbone while orchestrating a transition to IP-based, fiber-delivered services. That is a hard engineering posture under any conditions; it becomes brittle when the legacy estate is labeled unsustainable. The Government Accountability Office has repeatedly flagged FAA air-traffic-control systems for obsolescence risk and slow modernization pacing; in 2024, GAO reported that dozens of ATC systems were classified as unsustainable or potentially unsustainable, with modernization programs behind schedule on the most critical elements. Those assessments explain why a routine copper-to-fiber swap—mundane in a well-segmented enterprise—can still expose national-level fragility when performed inside an aging, tightly coupled architecture.
The lesson is not to pause modernization—quite the opposite. It is to execute transitions with the rigorous safeguards critical infrastructure demands: dark-site staging, pre-change soak tests, ironclad rollback plans, rate-limiting and storm-control at choke points, change windows aligned with systemic risk, and independent oversight on live cutovers. The FAA’s service restoration after isolating the misbehaving element demonstrates operational discipline; the scale of the initial blast radius argues for deeper architectural segmentation and more assertive fail-safes at the boundary between legacy and new.
Transportation Secretary Duffy sharply criticized contractor L3Harris over the FAA outage, saying there could have been mistakes during a router switch. He also said it’s unclear whether the cause was contractor error or faulty equipment.https://t.co/Vs33fKKw4m
— Jonathan Htet (@jonathan_htet) November 14, 2026
What this means going forward: accountability and architecture
Two responsibilities run in parallel. First, case accountability: the FAA’s root-cause analysis should settle whether the initiating condition was installation error, configuration error, or defective hardware—and ensure recurrence controls flow from that finding. L3Harris’s participation, already on record, should include work orders, method-of-procedure documents, configuration images, and device diagnostics preserved from the maintenance window. Second, architectural accountability: the agency’s modernization program must compress the window in which a misbehaving edge device can propagate instability facility-to-facility. That is design work—segmentation, rate limiting, isolation controls, independent monitoring—not merely vendor management.
Sources:
washingtontimes.com, l3harris.com, bloomberg.com, gpb.org, edition.cnn.com, floridatoday.com, foxla.com, finance.yahoo.com, kbzk.com, gao.gov



