Amazon Web Services suffered an outage on Tuesday that limited access to websites, business applications and streaming services including Netflix and Disney Plus for hours.
Network devices impaired by malfunctioning APIs on the cloud provider's East Coast data center in Virginia downed many websites and services. AWS products that went dark temporarily included the Chime video-conferencing service and the Connect contact center application.
Bloomberg reported that the outage disrupted Amazon's package delivery service by shutting down communications between the e-commerce company and its delivery drivers.
The outage struck in the morning, with website-testing site Downdetector receiving a flood of reports between 10:15 a.m. and 10:30 a.m. EST. Reports hit a peak at 11:20 a.m. and then tapered off. AWS continued to experience problems at 4:30 p.m., but the number of reports was about half of the peak.
The AWS status page reported that network devices affected by the API problems were the incident's root cause. The outage affected AWS' network monitoring and response tools, which delayed restoring service.
Most of the AWS platform was working by Tuesday afternoon, but the company did not know when it would fully recover. Amazon did not respond to a request for further comment.
The widespread impact of the outage results from an increasingly centralized market in which thousands of websites and services run on the leading cloud providers. According to research firm Statista, AWS accounted for 32% of the cloud market in October 2021, followed by Microsoft's Azure at 21% and Google Cloud at 8%. Other cloud providers held 39% of the market, down from 42% in the year's first quarter.
Gartner analyst Sid Nag said the largest cloud providers have become "too big to fail."
"The impact is much bigger than it was maybe five or six years ago," he said. "Given the criticality of the cloud in our day-to-day lives today, [outages] hit the front page of The New York Times and The Wall Street Journal."
Nag said many firms invest in multi-cloud strategies to ensure resilience. In those cases, applications can fail over to a different cloud provider when one provider goes down. However, that doesn't work for large-scale internet companies, Nag said.
"If you have a very hyperscale application running, you can't really fail it over to another cloud," he said. "That's where the real problem lies."
The outage comes one week after AWS' re: Invent conference, during which Nasdaq announced it would move its North American markets to the Amazon cloud.
Mike Gleason is a reporter covering unified communications and collaboration tools. He previously covered communities in the MetroWest region of Massachusetts for the Milford Daily News, Walpole Times, Sharon Advocate and Medfield Press. He has also worked for newspapers in central Massachusetts and southwestern Vermont and served as a local editor for Patch. He can be found on Twitter at @MGleason_TT.