Microsoft 365 outage, November 12–13: What happened, and what should you do?

In brief

On Thursday, November 12, users across large parts of the world had trouble opening their mailboxes in Exchange Online. The issue spread to a number of other Microsoft 365 services and continued into Friday. Microsoft is tracking the incident under EX1550543 and MO1550821.

On Friday afternoon, Microsoft reported sustained improvement, including in Europe. When this was written, at 6:00 p.m. Central European Time on Friday, November 13, the incident had not been declared resolved, and some users could still see intermittent access issues and delayed email.

Timeline

All times are Central European Time and come from Microsoft's own incident updates.

• November 12, 5:17 p.m.: The issue starts, according to Microsoft.
• 7:26 p.m.: Microsoft opens a separate incident for other affected services.
• 7:57 p.m.: Outlook traffic has been redistributed across regions.
• November 13, 2:06 a.m.: Microsoft narrows down the fault and rolls back a network configuration change.
• 8:25 a.m.: A fix improves stability.
• 10:55 a.m.: The delivery window for delayed email is extended to 48 hours.
• 2:00 p.m.: Sustained improvement across much of Europe.
• 3:01 p.m.: Email queues are significantly reduced and near normal levels.
• 4:53 p.m.: Microsoft is working on the remaining impact.

What was affected?

Mailbox access was affected whether users opened Outlook on a PC, on the web, or on a phone. Email also arrived with delays.

Because many services depend on Exchange Online, several others were affected: the Teams calendar, Copilot Chat, parts of SharePoint sharing, features in Defender and Purview, the admin center, and Intune. According to Microsoft, Teams calls and chat were largely unaffected.

What has Microsoft said about the cause?

Microsoft has not stated a final root cause. This is what the company has said so far:

• The load did not come from more traffic, but from many servers unexpectedly becoming unavailable.
• The servers were failing automated health checks.
• A process that establishes secure connections became overwhelmed, so the servers stopped responding.
• A recent network configuration change coincided with the start of the issue and was rolled back.

Microsoft has repeated in every update that there is no evidence of malicious activity. The company has promised a full report with root cause and corrective actions once the issue is fixed.

Was email lost?

Microsoft has not reported any lost email. To preserve delayed messages, Microsoft extended the delivery window to 48 hours. By Friday afternoon, the queues were near normal levels.

That means email sent during the incident will in most cases arrive, just later. Resending the message is rarely necessary.

How we followed up

We posted a service notice on the For customers page on Thursday evening and updated it when Microsoft reported improvement. The page also explains how customers reach us when something isn't working.

Where can you check the status yourself?

Microsoft publishes status in three places. Administrators see full details under Health and Service health in the admin center. The public status page also works when the admin center is down.

Microsoft's public status page

How to check Microsoft 365 service health

Microsoft's policy on incident notifications and post-incident reports

For the full sequence of events, Cloud Admin Hub has republished Microsoft's updates on EX1550543. MacRumors has summarized the scale with figures from Downdetector.