This weekly feature from S&P Global Market Intelligence, in collaboration with internet-service monitoring company ThousandEyes, aims to give remote workers insights into internet service disruptions.

Brief outages at two highly connected network transit providers drove overall outages among internet service providers up 128% week over week in early December.

Outages among U.S. ISPs rose to 139 for the seven days ended Dec. 11, from 61 the week prior. U.S. outages accounted for 52% of all global outages among ISPs, according to data from ThousandEyes, a network-monitoring service owned by Cisco Systems Inc. Global outages overall jumped 60% week over week, to 272 from 171.

However, many of those outages occurred outside of business hours, lessening the impact on remote workers. Thirty-six percent of U.S. outages in the past week occurred during business hours, compared to 40% the week prior, according to Angelique Medina, director of product marketing for ThousandEyes.

For the second consecutive week, there were no U.S. outages attributable to network connection problems among cloud-based collaboration apps, such as Zoom Video Communications Inc.'s video conferencing service or Microsoft Corp.'s Teams. There were two outages among collaboration applications worldwide. Notably, ThousandEyes' monitoring service only flags problems with networks connecting collaboration apps to internet users, not any internal issues that may cause problems but are not due to network errors.

Much of the increase in ISP outages related to trouble at network transit providers that connect local and regional service providers with the internet backbones and long-distance international connections that allow full access to the global internet. Outages among transit providers often cut off access to the ISPs they serve, dramatically increasing the impact of even brief network errors.

One such cascade effect was evident following a four-minute outage at Cogent Communications Holdings Inc., a U.S.-based multinational transit provider that experienced problems beginning at 4:50 p.m. PT on Dec. 8 on network nodes in the U.S., Germany, the U.K., Spain, France, Switzerland and Ireland, according to ThousandEyes' data.

The outage was cleared by 4:55 p.m. PT, but not before causing access problems for Cogent's downstream customers and for services hosted by Microsoft, Amazon.com Inc., SAP SE, The Walt Disney Co.'s streaming service and Wells Fargo & Co.

A 17-minute outage that began at 2:11 p.m. ET on Dec. 10 on an Atlanta-area network segment owned by network transit provider Hurricane Electric LLC of Fremont, Calif., caused a similar cascade of outages that affected the U.S., Canada, Germany, Egypt, Sweden, France and the U.K.

After five minutes, the outage expanded from Atlanta to affect Hurricane Electric network segments in Ashburn, Va.; Dallas; and the New York City area. The problem was resolved at 2:38 p.m. ET.

In the EMEA region, 41% of all outages occurred within business hours, while 40% of all outages occurred within business hours in the Asia-Pacific region.