The software king of the world, Microsoft, has brought large chunks of the world’s business communications to a standstill after its Microsoft 365 infrastructure suffered a spectacular failure.
According to Reuters, the outage, which began on 12 November, knocked out Outlook and Exchange Online before spreading to services including Teams, SharePoint, Copilot and parts of Microsoft’s security infrastructure. By Friday morning, the company was still struggling to persuade its servers to do the jobs customers were paying them to do.
At its peak, outage tracker Downdetector recorded more than 130,000 reports of Outlook problems in the US. The disruption then spread across Asia-Pacific, leaving businesses and government departments to discover how much of their daily operations depend on Microsoft’s computers behaving themselves.
Among those affected were the Australian Federal Police and domestic intelligence agency ASIO. Japanese government systems and the US House of Representatives were caught up in the disruption, which is hardly the sort of customer list Microsoft wants appearing in an incident report.
The problems began in Exchange Online, Microsoft’s cloud-based email and calendar service. Customers found themselves unable to access mailboxes through Outlook’s web interface, desktop software or mobile applications, while messages became delayed or failed to arrive.
Vole initially blamed elevated server loads caused by an unexpectedly large number of machines becoming unavailable. The remaining servers were apparently expected to pick up the slack, which worked about as well as asking the survivors of a shipwreck to tow the vessel home.
Attempts to restore the affected infrastructure produced disappointing results. Some servers returned to service following configuration changes, only to become unresponsive when customers started using them again.
A spokesVole said adjustments to its infrastructure had produced partial recovery in some regions, although the improvements initially failed to last. Engineers later investigated traffic-routing problems and increased server capacity to keep the remaining infrastructure running.
The disruption extended beyond email. Parts of Microsoft Teams lost calendar and scheduling functions, while services dependent on Exchange Online experienced problems. SharePoint, Microsoft 365 Copilot and components of Defender XDR were among the products affected by the wider incident.
The failure caused headaches for companies using Microsoft’s services as part of their products. OpenAI and Anthropic experienced problems with Microsoft integrations, while Indian airlines IndiGo and Akasa Air reported disruption to their email support operations.
Australian Home Affairs and Cyber Security Minister Tony Burke said Microsoft believed the outage involved a failed software update. The company has found no indication of malicious activity, although the precise technical cause remains unclear.
The scale of the disruption raises uncomfortable questions about the wisdom of allowing a handful of technology companies to control communications infrastructure used by governments and businesses worldwide. Microsoft’s cloud strategy has encouraged customers to surrender their local infrastructure in exchange for services that can apparently disappear across several continents following a configuration failure.
By Friday morning, Vole began reporting improvements after applying a patch to an authentication and routing component. The company said server stability was improving and engineers were extending the changes to additional regions. Microsoft has promised a post-incident report explaining the root cause and the measures intended to prevent a repeat. In the meantime, its engineers continued restoring services, with no confirmed timetable for full recovery.
As far as I am concerned, it has been borked on and off for more than 24 hours, and I am not best pleased.
