top of page
Black lettering reading "GP" on a yellow background.

Microsoft outage frustrates communities worldwide

Avery Slavin
2 minutes ago
3 min read

By Avery Slavin News Editor

Meghan Petrone / THE GATEPOST
Meghan Petrone / THE GATEPOST

On the afternoon of Nov. 12, senior Max Flecchia was waiting for a response to his email regarding an important career move. “I didn’t know how long it would take. I was being held up from potentially scoring a job interview,” he said. He was not alone. Multiple students and faculty members were waiting on emails all day, and unable to send any themselves. The thing holding them up? A worldwide shutdown of Microsoft. Multiple services powered by Microsoft 365, including Teams, Copilot, and Exchange Online, were inaccessible and not working properly. Exchange Online is the cloud-based background service that manages and stores the emails, contacts, and calendars associated with Outlook, meaning people were unable to send or receive emails as usual. Junior Brigid Milligan said while her schoolwork wasn’t necessarily affected, her job at an animal hospital was, as the staff was unable to obtain records for some pets. Microsoft’s preliminary post-incident report (PIR) stated the cause was an old Windows server defect. Increased traffic and reduced infrastructure capacity were also responsible for the disruption. Microsoft’s investigation of the outage uncovered that the previously undetected defect was introduced through a Windows update to the network in July of this year. However, it had gone undetected due to the fact that the system tests were being run in a way that would not cause the issue to arise, according to an article published by data analyst Sebastian Castellanos. The PIR said the defect was the internal Windows process experienced lock contention, which occurs when certain sessions need the same protected database at the same time and begin to wait on one another. This caused the central processing unit and the memory to become “exhausted,” making the servers’ response times slower or even completely unresponsive. Another factor of the Microsoft outage was the digital traffic was automatically redirected to the primary gateway that is shared with Exchange Online as opposed to an alternative networking gateway, according to the PIR. One final factor Microsoft said contributed to the outage was the server capacity being 14% unavailable from an unrelated, unspecified issue from the previous day. The increased workload from these problems on an already affected infrastructure triggered over 80% of servers to stop responding within two hours. Microsoft said the immediate recovery was difficult because the servers continued to shut down repeatedly although they were being restarted. The PIR posted Nov. 15 said the outage began at approximately 11:20 a.m. EST on Nov. 12 and concluded at approximately 11:10 a.m. EST the following day, or almost 24 hours later. Once the problem was fixed, Microsoft said the company added additional safeguards for traffic distribution, increased capacity, and expanded infrastructure monitoring. Microsoft said the outage was technical and the company did not suspect any malicious intent was involved. Patrick Laughran, associate vice president and chief information officer, said FSU “uses multiple layers of cybersecurity protection, including multifactor authentication, security monitoring, access controls, security awareness, and incident response planning. No organization can eliminate cyber risk completely, so we focus on both prevention and being prepared to respond and recover when something happens.” Laughran said Framingham State “has backup, recovery, and continuity measures in place for important University systems and information. However, if a cloud provider such as Microsoft has an outage, having backups does not necessarily make that service immediately available. “The recent incident with Microsoft is a good reminder for students to keep a second copy of important work on their personal computer rather than relying entirely on any single cloud service,” he added. Laughran said members of the community can check the FSU client portal home for technology service outage information in the case email is unavailable, and that the University may use additional communication methods during an emergency, such as FSU alerts.

  • Instagram
  • Facebook
  • Twitter
bottom of page