On the afternoon of November 13, 2026, the Federal Aviation Administration (FAA) ordered a nationwide ground stop across the United States. For more than four hours, planes stayed on the ground and thousands of flights were delayed or cancelled. There was no storm and no cyberattack: the cause was a network equipment change.

What Happened

According to information published since, a contractor was replacing a router at an FAA telecommunications site in El Dorado, Arkansas. The new device began flooding the FAA's telecommunications network with connection attempts. The overload spread until it destabilized most of the country's en-route air traffic control centers.

The most telling part is that the work was planned and the equipment had passed initial testing. The problem appeared a few minutes after it went live. The FAA has since paused installation of these routers and is reviewing its rules, which allowed this kind of maintenance during the day at remote sites.

The Real Problem: A Poorly Controlled Change

A faulty router happens. What turns a local incident into a national outage is the lack of guardrails around the change:

  • Timing — Doing the work during peak hours leaves no margin if something goes wrong.
  • Tests that are too short — A test that passes at go-live does not prove the equipment will behave under real load ten minutes later.
  • Monitoring — If no one is watching the network right after the change, the problem spreads before it is detected.
  • Rollback — Without a clear, fast way back, every minute spent diagnosing is expensive.

What It Means for an SMB

You do not have 22 control centers, but you do have a firewall, switches, a server and a Microsoft 365 subscription. A firewall change on a Tuesday at 10 a.m., a server update launched "quickly", or a new switch plugged in without checks can stop your whole business for a day.

Here are the simple rules we apply for our clients:

  • A maintenance window — Changes that affect the network or servers happen in the evening or on weekends, unless it is an emergency.
  • A written plan, even a short one — What, why, who, when, and above all how to roll back: saved configuration, old device kept close at hand.
  • An observation period — After the change, we monitor the network for at least 30 minutes before calling the work done.
  • One change at a time — If three things change at once and an outage occurs, there is no way to know which one caused it.
  • Contractors too — A vendor working on your network must follow the same rules as your own team.

The Bottom Line

The FAA outage is not a story about failing technology: it is a story about process. The good news is that change management requires no big budget or expensive software, only discipline. If you are not sure who changes what on your network, and when, now is a good time to talk to us.