However, after more than a day of missed meetings, backlogged emails, and nervous IT departments, Microsoft has stated that the worst is behind them.
Microsoft declared on Saturday, November 14, that its services had been restored following the major failure that started on Thursday.
"We have finished addressing the root cause of the incident and restoring service to all remaining affected customers and regions," the company said.
An explanation from the company has been provided. The explanation stated that the problem "was caused by an issue from a previous network capacity upgrade that came up after a recent network update."
According to Microsoft, the incident is unrelated to a cyberattack and to the activities of AI agents. This comes as a result of what the company has previously mentioned regarding the server problems as being behind the incident, including the Copilot AI assistant, and no malicious activity was observed.
The explanation is solely from Microsoft, and there have been no independent technical analysis or post-incident report in the reviewed reports.
The power failure was worldwide. The outage was reported in 74 different countries, with the outage affecting Outlook, Teams, and Copilot and having an effect on governments, courts, healthcare systems, corporations, airlines, and banks.
The system crash caused problems in healthcare operations, delayed earnings reporting, and made it difficult for schools to communicate.
Recovery did not happen overnight. On Friday, Microsoft claimed that services have been experiencing “broad recovery” by noon, but it also announced that it does not have any schedule for a potential resolution.
On Friday, email traffic was “approaching normal levels,” and Microsoft extended email timeout period to 48 hours due to the outage. Service access may restore before mail flow is completely back to normal according to Microsoft.
City of Jacksonville, Florida, said that it had some services restored in the morning and was “cautiously optimistic” about more success.
Microsoft issued an apology on Friday through X. “We are deeply sorry for the disruption this incident has caused our customers and partners,” the company wrote on X.
However, the company’s first statement about the problem was not as broad as the actual outage. First, it announced that it is investigating outage reports “in the Americas,” but it later found out that the outage had a global effect.
This rationale follows a familiar pattern. Any alteration to the infrastructure always leads to mass failure due to the fact that an error may lead to massive failures in other systems as well.
On July 2024, TechRadar reported that another Microsoft outage started when the change made to backend configuration settings caused an interruption between Azure storage clusters and the compute resources located in one region. This incident was a different one and on a smaller scale, but the initial catalyst remains the same.
None of the articles analyzed mentioned who sanctioned the current network upgrade or how the mistake remained undiscovered.
Questions arise. Why did a system capacity improvement lead to a global failure? Why did recovery take over a day? What changes will Microsoft introduce to prevent such problems in the future?
These questions have been left unanswered in the reports reviewed by me. This is because Microsoft normally issues a more thorough report about any major problem faced, and the said report would be relevant for contract considerations.
The repercussions may be lingering too, as there will still be late emails, and some users will have to verify whether meetings, files, and automatic processes were completed. Organizations that had deadlines during the downtime may have to account for this to their clients and/or regulatory bodies.
Companies might also assess their backup methods of communication. This is based on the accounts in the reports analyzed where offices and municipalities made use of other means of communication such as calling while the email system was down.
