Microsoft Azure cloud services restored after major outage; here’s what happened

Microsoft has confirmed that it has restored its Azure cloud services following a widespread outage that lasted several hours ahead of the company’s scheduled Q3 earnings announcement.

The disruption affected key platforms including Microsoft 365, Outlook, Xbox Live, and Copilot, temporarily disconnecting millions of users across various industries.

Microsoft services restored

Azure reported “strong signs of improvement across affected regions and are tracking toward full mitigation" by 7.20 pm ET Wednesday (October 29) (4.50 am IST Thursday). Affected services include Azure Communication Services and Media Services, among others.

Also Read: Is ChatGpt down? Outage hits users worldwide, over 500 outage reports

According to Microsoft’s Azure Status History, the outage was caused by an inadvertent tenant configuration change, which triggered a widespread service disruption across multiple regions. A rollback to a stable configuration was deployed, gradually restoring normal operations.

“After extended monitoring, we confirmed that the impact stemming from the Azure configuration change has been resolved,” Microsoft said.

“Engineers have confirmed that the issue impacting a subset of Azure services is now mitigated,” Azure support added on X.

Azure service outage

Microsoft acknowledged the outage on its Azure and 365 support accounts on X, confirming that users were facing access issues across multiple services.

“We’re investigating an issue impacting several Azure services,” the Azure support account stated. "Customers may experience issues when accessing services,” it noted.

Between 15:45 UTC (Coordinated Universal Time) (9.15 pm IST) on October 29 and 00:05 UTC (5.35 am IST) on October 30, customers leveraging Azure Front Door (AFD) may have experienced latencies, timeouts, and errors, Azure status report said.

Affected services included App Service, Azure Active Directory B2C, Azure Databricks, Azure SQL Database, Azure Virtual Desktop, Microsoft Sentinel, Microsoft Copilot for Security, and others.

Although most error rates and latencies have returned to normal, a small number of customers may still be seeing issues and we are still working to mitigate this long tail, the statement said.

Microsoft said that customer configuration changes to AFD remain temporarily blocked.

What went wrong and why

An inadvertent tenant configuration change within Azure Front Door (AFD) triggered a widespread service disruption affecting both Microsoft services and customer applications dependent on AFD for global content delivery, the Azure status report said.

The change introduced an invalid or inconsistent configuration state that caused a significant number of AFD nodes to fail to load properly, leading to increased latencies, timeouts, and connection errors for downstream services.

Also Read: Microsoft's largest layoff in years hits Xbox, sales, other divisions

As unhealthy nodes dropped out of the global pool, traffic distribution across healthy nodes became imbalanced, amplifying the impact and causing intermittent availability even for regions that were partially healthy.

The support team immediately blocked all further configuration changes to prevent additional propagation of the faulty state and began deploying a ‘last known good’ configuration across the global fleet.

Microsoft's protection mechanisms, to validate and block any erroneous deployments, also failed due to a software defect which allowed the deployment to bypass safety validations.

How the outage was resolved

Microsoft explained that it rerouted affected traffic to alternate healthy infrastructure as an immediate mitigation measure while investigating the root cause.

“We’ve identified portions of internal infrastructure that are experiencing connectivity issues. We’re unblocking these systems and redistributing traffic to support recovery,” the company said.

Engineers identified a recent configuration change to a portion of Azure’s infrastructure, which is believed to have caused the impact, and pursued several remediation strategies, including moving traffic away from the affected infrastructure and blocking the offending change.

The rollout of the faulty configuration was halted, and service traffic continued to be routed away from affected systems to restore availability.

The infrastructure was then reverted to a previously healthy state by deploying a stable configuration to the impacted systems. This was carried out alongside efforts to rebalance traffic across healthy infrastructure to mitigate the impact swiftly.

Microsoft confirmed that service health had largely recovered, although user reports of issues had not yet returned to pre-incident levels. Extended monitoring was implemented to ensure ongoing stability.

Strong earnings amid service disruptions

Despite the major outage, Microsoft reported robust Q3 earnings. However, users on social media reported difficulties accessing Microsoft websites and services, with Downdetector data showing issues beginning around 11:40 am ET.

Also Read: Microsoft layoffs: Xbox division likely to face largscale jobcuts

The Azure outage’s impact was felt globally: Alaska Airlines experienced disruptions to its website and booking systems, while Heathrow Airport and Vodafone UK also reported temporary service interruptions.

The Azure incident came just days after a similar global outage at Amazon AWS, which had affected thousands of websites and popular apps such as Snapchat and Reddit.