Microsoft said it’s making progress as it restores services following Thursday’s global outage.

The tech giant said “unexpectedly unavailable” servers impacted their infrastructure, triggering outages that impacted millions of people. Outlook, Teams, and other Microsoft services completely went down, but the email service is getting close to “normal levels” on Friday.

Isak Nti Asare is Assistant Dean for Undergraduate Education and Student Affairs at the Hamilton Lugar School, Co-Director of the Cybersecurity and Global Policy Program, Executive Director of the IU Cybersecurity Clinic, and a Fellow of the Indiana University Center for Applied Cyber Security Research. He said the problem wasn’t that too many people were trying to use Outlook at the same time.

“It was rather that Microsoft infrastructure couldn’t properly handle the traffic it was already receiving,” he said. “They became overwhelmed and stopped responding.”

Microsoft traced the problem to its cloud infrastructure, where servers are responsible for handling connections. Asare compared Microsoft 365 to a shopping mall.

“Every store, every app is different, but they share the same underlying infrastructure,” Asare explained.

Whether an outage to this scale can happen again with one of the world’s largest companies, Asare said there will certainly be disruptions as long as the demand curve is growing faster than the supply curve can. This brings data centers into the conversation, where Asare goes back to how reliant society is becoming on single digital infrastructure.

“All of the usage that we’re all now dependent on in our day-to-day lives is requiring a more robust digital infrastructure,” said Asare.

For companies like Microsoft, Asare said they don’t set out to build a system that never fails. Their goal instead is to build a system that fails gracefully and recovers quickly, where the cascading effects aren’t so great either.