Two outages in two days: what Microsoft 365 and the FAA ground stop teach about email backup

ARA Industries / Continuity · Facts checked 14 November 2026, 9:02 AM (PHT).

In the space of two days, two very different systems went quiet. First, a Microsoft 365 outage left businesses around the world without reliable email. Then a telecommunications fault led the US Federal Aviation Administration (FAA) to stop every departing flight in the country.

Officials say the two events are not connected. They still teach the same lesson: when one system carries all of your communication, you need to decide in advance what happens on the day it stops.

For a live, hour-by-hour account of the Microsoft incident and what to do while it is happening, read our Microsoft 365 outage guide. This piece looks at the bigger picture, and at the backup plan we suggest once the dust settles.

What happened with Microsoft 365?

The outage began on Thursday 12 November, US time, in the early hours of Friday 13 November in Manila. It started in Exchange Online, Microsoft's business email service, and spread to features that depend on it. Microsoft's status notes listed calendar-related failures in Teams, Copilot experiences that could not reach email, calendar and files, and SharePoint features that depend on Exchange Online. [1]

Microsoft said a process on affected servers "responsible for establishing secure connections becomes overwhelmed, causing the servers to stop processing requests." On Friday morning, US time (late Friday night in Manila), it said email queues were "significantly reduced and approaching normal levels." [2] Microsoft also said it had no evidence of malicious activity, and that it would publish a full post-incident report with the root cause. [3]

Status at our check (9:02 AM PHT, 14 November): Microsoft's public status feed still listed the incident as service degradation. Its latest note, from 6:00 AM (PHT) and still current at 9:02 AM (PHT), described "sustained service recovery," with Microsoft monitoring for an extended period and making sure the backlog of queued email processes before it calls the issue resolved. [1]

What happened with the FAA ground stop?

At 3:54 PM Eastern time on Friday 13 November (4:54 AM on Saturday 14 November in Manila), the FAA ordered a national ground stop "due to extraordinary circumstances and for reasons of safety." All departing traffic was held, and inbound international flights were managed case by case. [4] The FAA linked it to communication issues affecting multiple Air Route Traffic Control Centers, the regional centres that handle planes between airports. [8]

It was only the second nationwide ground stop since the 11 September 2001 attacks; the other was in January 2023. [8][9] As enroute facilities recovered, the FAA began letting some flights depart in coordination with its command centre, [5] and San Francisco International cleared some Pacific-bound flights, including to Asia. [11][12] Delta and United offered to waive change fees for affected travellers. [10]

The disruption was large. By 4:00 PM Pacific time on Friday (8:00 AM on Saturday in Manila), FlightAware counted about 3,600 cancellations and more than 10,200 delays across the US, according to the Los Angeles Times. [18] Those counts were still rising.

Status at our check (9:02 AM PHT, 14 November): the FAA's operations plan, issued at 8:37 AM (PHT), said the national ground stop "remains in effect for a large portion of the country," while some coastal facilities had been able to release some international departures. It listed the nationwide ground stop, with selective releases, until 04:00 UTC (12:00 noon PHT). [5][17] BBC News, reporting at about the same time, said the stop had lasted more than three hours, longer than the 2023 event. [11]

Are the two outages linked?

Officially, no. Homeland Security Secretary Markwayne Mullin told reporters: "I don't see any reason to connect those two at all. Those are separate issues." Microsoft said it had no indication that the ground stop was related to its service disruption. [6][7]

Some people online also asked about SMART, the FAA's new AI-based traffic-planning tool, which had just been expanded to all 22 en route centres. Asked about a possible link, Mullin said, "We're looking at that." The company behind SMART said the ground stop "is not based on or a result of the SMART system." [6][7] Mullin said there was no indication of an attack, while adding that officials were "not ruling out anything at this point." [6][9]

The cause had not been officially confirmed at our check. In an unconfirmed report, CBS News, citing multiple sources, said that an update to the private communications network linking US air traffic control caused it to reboot and then crash. [11] This has not been confirmed by the FAA; treat it as a report, not a finding.

Our view: two big outages close together will always invite theories. The useful question for a business is not whether they were linked, but what it would do if its own main channel went quiet for a day.

The shared lesson: single points of failure in communication

Both incidents show what happens when one system carries everything.

• Plan your "ground stop" before you need it. The FAA's choice was to stop new departures and keep flights already in the air safe. Decide in advance what your team pauses (new quotes, payments, client releases) and what carries on when email is down, and who makes that call.
• Recovery comes in stages. Microsoft brought servers back gradually and extended its email timeout to 48 hours to protect queued mail, [3] and the FAA released flights as each centre recovered. Expect your own recovery to be partial for a while, and tell staff and clients that plainly.
• Keep a second way to reach people. If email and calendars fail together, an SMS list, a chat group or a simple status page keeps staff and key clients informed.
• Watch for copycat scams. Outages bring fake "fix your mailbox" and "rebook your flight" messages. Remind staff to check news only through official channels.

A common question after the outage: "What should our backup plan be?"

After the Microsoft outage, a common question from businesses we work with was simple: what should our backup plan be? Here is the short version of our advice.

1. Hold and queue mail with a continuity service

A third-party mail continuity service sits alongside your main provider. When Microsoft 365 is down, it can accept and hold your incoming mail, and give staff a basic web inbox to read and reply to urgent messages. When service returns, it hands the mail back. This is usually the least disruptive option for a smaller team, because nothing changes on a normal day.

2. Set up MX failover in advance, with SPF and DKIM first

You can also prepare a DNS switch, changing your MX records to send mail to a backup route, such as a Google Workspace mailbox, during a long outage. Two things matter:

• Senders retry. Mail servers normally keep retrying delivery for days. The email standard suggests a give-up time of at least four to five days, [13] so a short outage rarely loses mail. A switch is for long outages, not every hiccup.
• Authentication has to be ready. If your backup route sends mail, publish SPF for it and turn on DKIM before you move DMARC to quarantine. Google advises setting up DKIM and SPF at least 48 hours before DMARC and rolling the policy out gradually. [14] Under a quarantine policy, mail that fails DMARC can be sent to spam, [15] so a half-finished backup can hide your messages just when they matter most.

Lower your DNS record lifetimes ahead of time, write down who can make the change, and test the switch on a quiet day.

3. Check how many of your contacts also run on Microsoft

A backup inbox helps only if the people you write to can still send and receive. Look at your most important clients and suppliers. If most of their domains point to Microsoft 365 (their MX records end in mail.protection.outlook.com), [16] they were down when you were. In that case your effort may be better spent on a non-email fallback, such as an SMS list or chat group, and clear client updates, than on a second mailbox.

Sources

1. Microsoft, Microsoft 365 public status feed (status.cloud.microsoft/api/feed/mac), item dated 14 November 2026, 6:00 AM (PHT), checked 9:02 AM (PHT). Status page: status.cloud.microsoft/m365.
2. ABC7 Los Angeles, "Microsoft outage live updates: What to know as service is being restored," updated 13 November 2026. abc7.com
3. University of Wisconsin–Milwaukee, Microsoft service outage notice quoting Microsoft's incident updates, 13 November 2026. uwm.edu
4. FAA Air Traffic Control System Command Center, Advisory 111, national ground stop, 13 November 2026. fly.faa.gov
5. FAA Air Traffic Control System Command Center, Advisory 120, nationwide ground stop update, 13 November 2026, 22:46 UTC. fly.faa.gov
6. GeekWire, "DHS chief sees no Microsoft link to FAA ground stop, says possible AI role being investigated," 13 November 2026. geekwire.com
7. Business Insider, "FAA SMART AI system vendor and Microsoft both say ground stop isn't their fault," 13 November 2026. businessinsider.com
8. Northeastern Global News, "Telecommunication issue grounds all U.S. flights one day after global Microsoft outage," 13 November 2026. news.northeastern.edu
9. The Washington Post, "FAA grounds flights at U.S. airports, citing 'telecommunications outage'," 13 November 2026. washingtonpost.com
10. ABC News, "FAA ground stop live updates," 13 November 2026. abcnews.com
11. BBC News, live updates, "Flights grounded across US as homeland security secretary says no indication of an attack," 13–14 November 2026. bbc.com
12. The San Francisco Standard, "Passengers cheer as SFO resumes some international flights amid FAA ground stop," 13 November 2026. sfstandard.com
13. IETF, RFC 5321, Simple Mail Transfer Protocol, section 4.5.4.1. rfc-editor.org
14. Google Workspace Admin Help, "Set up DMARC." support.google.com
15. IETF, RFC 7489, Domain-based Message Authentication, Reporting, and Conformance (DMARC), section 6.3. rfc-editor.org
16. Microsoft Learn, "External Domain Name System records for Microsoft 365." learn.microsoft.com
17. FAA Air Traffic Control System Command Center, Advisory 001, operations plan, 14 November 2026, 00:37 UTC. fly.faa.gov
18. Los Angeles Times, "LAX, other airports fill with stranded passengers," updated 13 November 2026, 4:15 PM PT. latimes.com

Also see the AP's live coverage of the ground stop: apnews.com. Sources checked 14 November 2026, 9:02 AM (PHT). Both stories are still developing; prefer official status pages for the live picture.