Global Network Outages Fall 25% as Microsoft and Lumen Report Brief Disruptions

ThousandEyes, the Cisco-owned internet monitoring firm, recorded 450 global network outage events during the week of November 2-8, a 25% drop from the 601 outages logged the previous week. The decline was driven largely by fewer public cloud network disruptions, even as ISP outages in the United States rose sharply, and two short but wide-reaching incidents hit Microsoft and Lumen networks.

What the weekly numbers show

Of the 450 outages tracked globally between November 2 and November 8, 234 occurred within the United States, an 8% decrease from 253 the week before. ISP outages worldwide ticked up 1% to 148, but in the U.S. they jumped 30%, from 57 to 74 events. Public cloud network outages moved in the opposite direction, falling 21% globally to 137 and 22% in the U.S. to 117. Collaboration app network outages stayed flat at zero for both the global and U.S. totals, continuing a pattern seen in several recent weeks. ThousandEyes compiles these figures by continuously testing how internet service providers, cloud platforms, collaboration tools and edge services such as DNS and content delivery networks are performing, flagging periods when connectivity degrades or fails outright.

Microsoft's multi-region disruption

On November 3, Microsoft experienced a network outage affecting downstream partners and services across the U.S., Canada, Spain, Switzerland, Germany, India and France. First detected around 3:02 p.m. EST near Dulles Town Centre, Virginia, the disruption spread roughly six minutes later to nodes in Chicago, and then to Baltimore as well, a pattern that coincided with a rise in affected customers. The incident lasted a total of 12 minutes across a 23-minute window and cleared by about 3:25 p.m. EST. The shifting locations of affected nodes illustrate how outages can migrate across a provider's infrastructure before full resolution, even when the total active-outage time is brief.

Lumen outage adds to carrier-level incidents

Two days later, on November 5, U.S. carrier Lumen (formerly CenturyLink) reported a 10-minute outage over a 17-minute span that began around 4:08 p.m. EST, centered initially on New York nodes before spreading to New Jersey. The disruption affected customers and partners in the U.S., India and Canada before clearing around 4:25 p.m. EST. Both incidents fit a recurring pattern ThousandEyes has documented throughout the year: short-duration outages at major carriers and cloud providers that nonetheless ripple outward to multiple countries because of how traffic is routed through shared backbone infrastructure.

Context from recent weeks

The November 2-8 period followed a week (October 26-November 1) in which outages rose 31% to 601 events, driven by a 24% increase in public cloud network disruptions. Public cloud outages have fluctuated considerably across the fall, reflecting how sensitive that category is to configuration changes, maintenance windows and routing shifts at major providers. ThousandEyes' methodology does not establish root cause for every event; the specific incident summaries, including Microsoft's and Lumen's, rely on observed network behavior such as node location and duration rather than vendor-confirmed explanations in every case.