‘I hope this email finds you’: Microsoft declares outage over as the world digs out of its inbox
Microsoft says it has restored Microsoft 365 service after an outage cut off email for governments, hospitals, courts and businesses in more than 70 countries. The company traced the problem to an earlier network capacity upgrade and says it will publish a preliminary post-incident report within two business days. Read More
Steven Sinofsky was relieved this week to finally give up a title he never wanted.
“Somewhat happy to say I am no longer responsible for the worst Outlook outage,” the former Microsoft executive posted on X Friday, sharing a May 2000 news story about the “I Love You” worm, which overwhelmed corporate email systems when he ran the Office group.
Later in the day, he aired a pet peeve. After a plumber emailed to say he’d missed an appointment because of the “Outlook outage,” Sinofsky noted that this one wasn’t really the application’s fault. “This time it is the data center layer,” he wrote.
Microsoft’s explanation backs him up. The company says the failure, which cut off email for people and organizations in more than 70 countries this week, was introduced during an earlier network capacity upgrade in its Microsoft 365 cloud and surfaced after a more recent network change. It hasn’t given more details.
The company said late Friday it had restored service to the remaining customers and regions. The disruption lasted about 30 hours, according to Microsoft’s incident log. Microsoft offered a public apology and promised a detailed post-mortem explaining what went wrong and next steps.
Charles Lamanna, Microsoft’s executive vice president of Copilot, Agents and Platform, said on X Friday afternoon that “cascading failures + thundering herds at our network perimeter” made recovery more difficult than normal.
A “thundering herd” isn’t a stampede of frustrated IT admins descending on Redmond, as fitting as that might have been. It happens when huge numbers of devices try to reconnect at the same time, overwhelming systems as they come back online.
He said the post-mortem would come “shortly,” with a “step-by-step timeline and our many repair items to avoid similar outages in the future.” In a reply to another user, he said Microsoft “will be very specific on what happened.” In its notice to customers, Microsoft said it would provide a preliminary post-incident report within two business days and a final report within five.
In the meantime, like prior incidents involving AWS and others, the “Outlook outage” became the latest example of how much depends on one company’s cloud.
Microsoft reported in January that it had more than 450 million paid commercial Microsoft 365 seats, roughly one for every 18 people on Earth.
Among those affected this week were Unity Health Toronto, a hospital network that had to declare a Code Grey, the alert for a systems failure; a Colorado jail that couldn’t take online bond payments; and Pennsylvania lawmakers who went about 18 hours without email. Unity Health said patient care continued without disruption.
In Japan, securities firms and asset managers fell back on fax machines when trade-confirmation emails stopped coming through, Business Upturn reported, citing Nikkei.
In New York, the city’s Emergency Management office warned of delays reaching city agencies. In Singapore, the outage disrupted some government systems, including email, and in London, the UK Supreme Court wasn’t receiving email.
About the only thing that wasn’t Microsoft’s fault was Friday’s nationwide FAA ground stop. Federal officials blamed aging FAA telecom gear. Homeland Security Secretary Markwayne Mullin opened an evening news conference — before even getting to the news that U.S. airspace had reopened — by saying there was “no connection between this and Microsoft,” adding, “I know a lot of people are making that jump.”
Microsoft itself looked to preempt the speculation a few hours earlier, posting on X that it had “no indication that the FAA’s nationwide ground stop is related to any Microsoft services.”
The Microsoft 365 outage was at least the fourth significant one this year. A January outage lasted nearly 10 hours, and in July a bug in Microsoft’s automated network maintenance system disrupted Azure and Microsoft 365 services for hours. On Aug. 31, an Exchange Online outage also hit Teams, OneDrive and SharePoint.
Instead of “I hope this email finds you well,” many people joked, it should just be “I hope this email finds you.” (Consider this our nomination for the “Overworked Twitter Joke of the Week” on The Ringer’s Press Box podcast.)
Microsoft said Friday afternoon that it was watching the backlog of delayed email “to ensure all process successfully before considering the issue resolved.” The University of Washington’s IT department told its users that email delayed during the outage would be delivered as mail queues continued to be processed.
Many people coped with the flood of returning email with a bit of levity.
As the mail started flowing again, CNN chief media analyst Brian Stelter said he was getting “a deluge of emails from PR people pitching their clients to talk about the Outlook outage.” Most were sent during the outage, he noted, “which, fittingly, prevented the pitches from reaching me.”
The security company Malwarebytes captured the paradox in a message to its followers: “We regret to inform you that Teams and Outlook are coming back online after a global outage.”
