July 23, 2026 Microsoft Azure and Microsoft 365 Microsoft experienced a major technical disruption to its services. A system outage during routine fiber maintenance in the US West Coast (US West / California) region affected Microsoft’s network connectivity for about five hours.
This network outage affected many global users, including the US, due to which Outlook, Teams, SharePoint, Copilot And Azure Sentinel Major cloud services were disrupted.
Timeline of the outage
Released by MicrosoftPreliminary Post Incident Review (PIR) According to the report, the complete timeline of the outage was as follows:
| Time (UTC) | situation and action |
| 14:44 UTC | Network device maintenance began and immediately users began experiencing connectivity issues. |
| 14:45 UTC | Microsoft’s networking and incident response teams began investigating the traffic anomaly. |
| 15:00 – 16:00 UTC | Engineers analyzed telemetry data to assess the blast radius of the problem. |
| 16:00 – 17:45 UTC | The cause of the error was found to be recent network maintenance and incorrect IP routing. |
| 17:45 UTC | The process of rolling back the changes made during maintenance was started. |
| 18:26 UTC | The rollback completed and network stability began to be restored. |
| 19:41 UTC | All Azure and Microsoft 365 services were fully recovered after approximately 4 hours and 57 minutes. |
What was the reason behind the error?
According to Microsoft’s initial assessment, the problem was not caused by an external cyber attack or a physical cable cut, but rather by a Software bugs in automated maintenance processes has been caused by
Isolation Process:Certain network paths had to be isolated for regular maintenance.
Request Converter Bug: The maintenance system’s automated request converter had a bug. It passed system checks, but exceeded the set limits for equipment and equipment during maintenance. IP Routes was removed from the network.
Traffic blackout: Although local traffic within the data center continued, all data networks outside of US Western were cut off.
Which services were affected?
The outage affected over 27 Microsoft services:
- Microsoft 365 Apps: Outlook, Teams, OneDrive, SharePoint, और Copilot AI।
- Azure Infrastructure: Azure Sentinel, Log Analytics, Azure Monitor, Application Insights, Azure Kubernetes Service (AKS), और Azure Firewall।
- Consumer Services: Xbox Live और Microsoft Store।
Only traffic that was going into or coming from outside the California/US West region was affected. Workloads running entirely within the region were not directly impacted.
Big lessons for tech companies and businesses
This incident shows that no matter how advanced the automation, a small coding mistake or routing bug can impact the entire global cloud infrastructure.
- Multi-region redundancy:Microsoft itself recommends that organizations handling mission-critical data should always adopt a multi-region architecture instead of relying on a single-region one.
- Automation Safety Checks Review:Continuous testing of automated scripts and security measures of AI-powered operations (AIOps) is essential.
- Faster Disaster Recovery:The ability to roll back immediately upon a maintenance failure helps minimize damage.
Microsoft has clarified that they are improving their automated maintenance validation system to prevent such incorrect IP routing from happening again in the future.
