Newshot

Microsoft Publishes Final Report on Global Outlook Outage: Some Emails Failed to Deliver, Security Features Also Affected

AI summary

Microsoft has released its final incident report on the November 12 global outage that took down Outlook, Exchange Online and other Microsoft 365 services for millions of users, disclosing that some emails were never delivered to their intended recipients and that several email security features were affected. The company said some delayed messages sent to personal Outlook mailboxes that had been inactive for more than 30 days expired and senders received non-delivery notices, while all queued enterprise email was eventually delivered. Microsoft also confirmed that Microsoft Defender for Office features were impacted, including delays in Zero-hour Auto Purge (ZAP) and issues with Safe Links, and it published improvement plans to prevent a recurrence.

Translation(model translation, not the source's own text)

Microsoft Publishes Final Report on Global Outlook Outage: Some Emails Failed to Deliver, Security Features Also Affected

IT之家 November 22 news: On November 12, Microsoft suffered a large-scale service failure that prevented millions of users worldwide from properly accessing Outlook, Exchange Online, and several other Microsoft 365 services. It took Microsoft about 12 hours to gradually restore all services. Last week, the company released a preliminary incident report explaining the cause of the failure.

Now, Microsoft has published the final incident report, disclosing details that were not previously made public: some emails failed to reach their intended recipients. In addition, the outage also affected several of Microsoft's email security features, and Microsoft has also announced improvement plans to prevent similar incidents from happening again.

IT之家 notes that last week's investigation showed the failure originated from a software defect introduced in July 2026. This flaw, combined with factors such as reduced server capacity and increased Microsoft Teams traffic, ultimately triggered a cascading failure that at one point left more than 80% of Microsoft's network gateway servers unable to respond properly.

During the outage, Exchange Online could not process emails properly, and a large volume of emails therefore piled up awaiting redelivery by the system. To prevent delayed emails from being returned because they exceeded the retry deadline, Microsoft extended the email delivery retry period from the original 24 hours to 48 hours.

Microsoft confirmed that all queued enterprise emails were eventually delivered successfully, and no permanent loss of enterprise emails was found. However, the situation differed for Outlook mailboxes for individual users.

Microsoft acknowledged that some delayed emails originally destined for long-inactive personal mailboxes had expired after the mailboxes remained inactive for more than 30 consecutive days, and senders therefore received notifications that the emails could not be delivered. Microsoft did not disclose the specific number of affected emails, nor did it say how many users received such notifications.

Email security features also affected

The final incident report also disclosed that the outage affected several important email security features of Microsoft Defender for Office.

For example, the Zero-hour Auto Purge (ZAP) feature experienced delays. This feature normally automatically removes malicious emails after they are delivered, but because the system could not obtain updated threat intelligence in a timely manner, purge operations were not carried out at the normal speed. This means some affected users may have been exposed to malicious email threats for longer than usual.

Microsoft also confirmed that users encountered problems when opening URLs protected by Safe Links. Safe Links performs a security check on a link before the user visits it, but depending on the organization's security policy, some users may see a "skip verification" option during the check.

At the same time, security administrators also encountered multiple operational obstacles, including releasing or deleting quarantined emails, managing allow lists and block lists, and performing email security remediation operations.

The impact of the outage also extended to Microsoft Purview. Policy changes, security alerts, and data processing tasks on the platform could all be delayed, involving features such as data loss prevention (DLP), eDiscovery, insider risk management, and audit search.

Microsoft announces multiple improvement measures

To avoid similar large-scale failures from happening again, Microsoft plans to make adjustments in multiple areas, including network traffic allocation, server capacity, and failure recovery mechanisms.

By the fourth quarter of 2026, Microsoft plans to revise its traffic allocation architecture and limit Teams failover to the affected region, preventing localized service abnormalities from further affecting other regions.

By the first quarter of 2027, Microsoft plans to divide network gateway server capacity into multiple independent resource pools according to different workloads, preventing problems in one service from affecting other services.

Microsoft also plans to expand the scope of load testing to include the latest generation of high-core-count servers, and to adopt traffic patterns closer to a real production environment so that potential problems can be discovered in advance.

Other improvement measures include: speeding up the recovery process after large-scale server failures, improving diagnostic capabilities for Windows system changes, increasing server capacity, narrowing the impact of regional failures, and preventing restored servers from collapsing again due to excessive load.

In addition, Microsoft acknowledged that the first outage notification to customers was sent only about 20 minutes after the large-scale failure occurred. To improve the user communication experience, the company plans to refine its automated outage notification mechanism by the end of 2026, striving to notify customers more promptly when service abnormalities occur.

Related articles

  • Some Microsoft classic Outlook users report shared mailboxes "vanishing into thin air" after a few seconds
  • Microsoft solicited AI tips during the Outlook and other service outage, and its tweet turned into a scene of netizen complaints
  • Microsoft says Outlook and other services are gradually recovering and stabilizing, while some users are still waiting for restoration
  • Microsoft executive responds to Outlook and other service disruptions: deeply regretful, working full-time on fixes
  • Fixed: Microsoft admits Outlook mistakenly blocked legitimate emails, rejected emails need to be resent
  • Search emails even without internet: Microsoft adds "local search" to classic Outlook on Win10/Win11

The translation is machine-generated and individual terms may be off. Nothing is rewritten or cut; the original is authoritative.

Published by IT之家
This site only summarises and aggregates; copyright stays with the original author. Headlines and summaries follow the source — where they differ, the original wins.
Read the original ↗