What Really Caused Last Week's Microsoft 365 Outage? A Timeline of What Went Wrong

On Thursday, November 12, millions of Microsoft 365 users began experiencing something businesses have come to dread: email stopped working, applications became inaccessible, and everyday operations were disrupted. What initially appeared to be an Outlook problem quickly became a much larger Microsoft 365 outage, affecting organizations around the world. Microsoft Teams, Exchange Online, SharePoint, OneDrive, Copilot, and several other services experienced varying degrees of disruption.
For businesses that depend on Microsoft 365 to communicate, collaborate, and serve customers, the outage was more than an inconvenience. It was a reminder of how dependent modern workplaces have become on cloud infrastructure. Microsoft has since released a preliminary report explaining what happened. And while the technical details are complicated, the underlying story is surprisingly straightforward.
Three separate infrastructure problems occurred at the same time, creating a domino effect that Microsoft's normal recovery systems couldn't stop.
Here's what happened, why it happened, and what businesses can learn from it.
What Actually Caused the Microsoft 365 Outage?
According to Microsoft's preliminary Post-Incident Report, released November 15, the outage resulted from three contributing factors.
1. A Hidden Software Defect That Had Been Present Since July
The first problem began months before anyone noticed an issue. In July 2026, Microsoft began deploying a software change to specialized Windows servers that manage connections to Microsoft 365 services. These servers act as gateways, processing requests from users attempting to access Outlook, Teams, and other cloud applications. The update appeared successful. It passed Microsoft's testing procedures and operated for more than three months without causing any noticeable problems. However, the software contained a hidden defect.
Under certain conditions, internal processes on the servers competed for the same resources, a technical problem known as lock contention. This could overwhelm the servers' processors and memory, eventually causing them to stop responding Interestingly, Microsoft's testing had not identified the issue because the problem was more likely to occur on newer, more powerful servers with additional processor cores. The software had worked normally for months. It simply hadn't encountered the combination of conditions that would trigger the defect.
2. Microsoft Teams Traffic Was Redirected to Shared Infrastructure
The second contributing factor occurred approximately 15 hours before the outage. Microsoft uses multiple gateway systems to distribute traffic and maintain service availability. When one system experiences problems, traffic can be redirected to another.
Early on November 12, Microsoft Teams traffic was automatically moved from one gateway system to another that also supported Exchange Online. This was a normal recovery operation, not an unusual event. However, the change increased the workload on the gateway infrastructure supporting Outlook and other Microsoft 365 services. The additional traffic also changed the types of requests those servers were processing. That increased workload made it more likely that the previously undetected software defect would be triggered.
3. Nearly 14% of Server Capacity Was Already Unavailable
The third factor was reduced infrastructure capacity. Microsoft reported that approximately 14% of the affected gateway system's server capacity was already unavailable because of an unrelated issue from the previous day. Under normal circumstances, the remaining infrastructure should have been able to handle the workload. But with additional Teams traffic arriving and the hidden software defect waiting to be triggered, the reduced capacity became a significant contributing factor. As servers began failing, Microsoft's automated systems redirected traffic to the servers that were still operating. Unfortunately, those servers were vulnerable to the same defect. More traffic caused additional servers to fail. Those failures pushed even more traffic onto the remaining servers.
In less than two hours, more than 80% of the affected gateway servers became unresponsive.
The safeguards intended to isolate failures between regions also failed to contain the problem, allowing the disruption to spread globally.
Microsoft 365 Outage Timeline: November 12–14, 2026
The following timeline is based on Microsoft's preliminary incident report. All times are Eastern Standard Time (EST).
| Date and Time | What Happened |
|---|---|
| Nov. 11 | An unrelated infrastructure issue left approximately 14% of gateway server capacity unavailable. |
| Nov. 11, 8:07 PM | Automated systems redirected Teams traffic to the gateway infrastructure shared with Exchange Online. |
| Nov. 12, 11:03 AM | Microsoft's monitoring systems detected increasing connectivity problems. |
| Nov. 12, 11:20 AM | Microsoft received a high-priority Exchange Online alert and began communicating the incident. |
| Nov. 12, 11:27 AM | Microsoft confirmed problems accessing mailboxes through Outlook and other connection methods. |
| Nov. 12, 1:27 PM | Microsoft expanded its incident notification to cover additional services, including Teams, Copilot, and Defender. |
| Nov. 12, 2:20 PM | Engineers determined that unexpectedly unavailable servers, rather than an overall increase in traffic, were overwhelming the remaining infrastructure. |
| Nov. 12, 5:00 PM | Initial recovery attempts were proving unsuccessful. |
| Nov. 12, 6:55 PM | Servers that had been restarted were becoming unresponsive again when traffic returned. |
| Nov. 12, 8:00 PM | Engineers identified an overwhelmed connection-handling process as a major source of the failures. |
| Nov. 13, 2:20 AM | Microsoft began deploying a fix to reverse the defective software behavior. |
| Nov. 13, 9:17 AM | Global webmail availability returned to expected levels. |
| Nov. 13, 11:10 AM | Microsoft confirmed that customers were no longer experiencing active impact and continued monitoring recovery. |
| Nov. 13, 11:15 PM | Microsoft issued its final customer updates after extended monitoring confirmed service stability. |
Microsoft's official incident window ran from 11:20 AM Thursday, November 12, through 11:10 AM Friday, November 13, nearly 24 hours. Some users experienced connectivity issues before the formal incident start, while monitoring and recovery of delayed messages continued afterward.
Why Was the Outage So Difficult to Fix?
One of the most concerning aspects of this incident was that Microsoft's normal recovery procedures initially made little progress. When a server stops responding, restarting it or moving traffic to another server is typically an effective solution. But in this case, the problem wasn't isolated to individual machines. Every time engineers brought servers back online, the accumulated traffic could overwhelm them again. Microsoft also couldn't simply reverse a recent software update.
The underlying defect had been introduced months earlier, meaning the company's standard rollback mechanisms weren't designed to quickly undo it. Engineers ultimately developed and deployed a targeted correction that restored the affected connection-handling software to its previously working behavior. Rather than restarting everything simultaneously, Microsoft restored groups of servers in a controlled process to prevent another wave of failures. That approach helped stabilize services and gradually restore access across regions.
Which Microsoft Services Were Affected?
Although Outlook received much of the attention, the outage affected a much broader collection of Microsoft 365 applications.
Exchange Online and Outlook
Email experienced the most significant disruption. Users reported being unable to access inboxes, send or receive messages, or connect through desktop, web, and mobile applications. Many messages were delayed while Microsoft's infrastructure recovered.
Microsoft Teams
Some Teams capabilities were disrupted, particularly features that depended on Exchange Online. However, Microsoft reported that Teams chats and meetings continued functioning for most users during much of the incident.
SharePoint and OneDrive
Certain features experienced connectivity problems, although these services were able to move away from the affected gateway infrastructure and recover sooner.
Microsoft 365 Copilot
Copilot features that relied on Exchange Online and other affected services experienced disruptions.
Microsoft Defender and Other Administrative Services
Some capabilities within Defender, Purview, Intune, Planner, Power BI, and other Microsoft services were also affected. Importantly, not every application was completely unavailable. The severity depended on the service, region, and connection method.
Was This a Cyberattack?
Given the scale of the disruption, many organizations understandably questioned whether Microsoft had experienced a cyberattack. Microsoft stated that it found no evidence of malicious activity associated with the incident. The preliminary findings point to an internal software defect combined with traffic changes and reduced infrastructure capacity. This distinction matters. Not every major technology outage is caused by ransomware, hackers, or another cybersecurity incident. Sometimes, the technology designed to keep services available can encounter unexpected failures of its own. And those failures can have consequences just as disruptive to daily business operations.
What Does This Mean for Businesses That Rely on Microsoft 365?
The outage highlights a challenge that organizations of every size face. Cloud platforms provide tremendous advantages, including scalability, accessibility, collaboration, and reduced infrastructure management. But moving critical business applications to the cloud doesn't eliminate operational risk. It changes where that risk exists.
For example, consider what happens when an organization suddenly loses access to its primary email platform. How does it communicate with customers? How do employees receive urgent instructions? Can departments continue working if calendars, documents, and collaboration tools are unavailable? These aren't questions businesses should be answering for the first time during an outage.
Five Lessons Businesses Should Take Away
1. Have a communication method that doesn't depend on Microsoft 365.
If Outlook and Teams are unavailable, employees still need a reliable way to communicate. Organizations should establish an alternative communication channel that operates independently of their primary cloud environment.
2. Understand which business processes depend on cloud applications.
Email is only one piece of the puzzle. Scheduling, customer communications, document access, reporting, and other essential functions may depend on interconnected cloud services. Identifying these dependencies helps organizations understand the potential impact of an outage before one occurs.
3. Review your business continuity plan.
Many organizations associate business continuity planning with natural disasters, power failures, or cybersecurity incidents. Cloud service disruptions deserve the same consideration. A good continuity plan should explain how essential operations will continue when a major technology provider becomes unavailable.
4. Make sure employees know what to do.
A backup communication platform is only useful if employees know how to access it. Businesses should document procedures, assign responsibilities, and periodically test their response plans.
5. Don't assume that cloud availability eliminates the need for IT planning.
Even when an outage originates entirely within a technology provider's infrastructure, internal IT teams and managed service providers still play an important role. They can help assess the impact, communicate accurate updates, identify available alternatives, and coordinate recovery once services return.
What Has Microsoft Done to Prevent This From Happening Again?
Microsoft reports that it has already taken several corrective actions. These include restoring the affected connection-handling code to its previously working state, increasing infrastructure capacity, improving monitoring, and introducing additional safeguards to prevent traffic redirection from triggering another cascading failure. Microsoft also extended its email delivery window to allow delayed messages to be processed rather than immediately returned as undeliverable. A final Post-Incident Report is expected to provide additional details about longer-term improvements and preventative measures. As of November 16, that final report has not yet been published.
The Bigger Picture: Business Continuity Is About More Than Backups
Last week's Microsoft outage demonstrated how quickly a problem within one technology platform can affect organizations worldwide. It also reinforced an important distinction: having your information stored in the cloud doesn't automatically mean your business can continue operating when cloud services become unavailable. Backups protect data. Business continuity planning protects operations. Both are essential.
At Total Communications, we help organizations build reliable IT environments that support their day-to-day operations and prepare them for unexpected disruptions. From managed IT services and network infrastructure to cybersecurity, unified communications, and business continuity and disaster recovery planning, our focus is helping businesses reduce risk and maintain productivity. Because while no organization can prevent every outage, having the right technology, processes, and support in place can make a significant difference in how well it responds.
Is your business prepared for the next major cloud outage?
Total Communications can help you evaluate your IT environment, identify potential vulnerabilities, and develop a continuity strategy that keeps your organization better prepared when critical services become unavailable.