Amazon Web Services (AWS) suffered an extended outage on Tuesday in its Eastern U.S. cloud region, the impacts of which rippled across a wide variety of consumer services.
In a notice posted to its service dashboard, AWS said the issue began around 12:30 pm ET (9:30 am PT) in its US-East-1 Region and identified the root cause as “an impairment of several network devices” which caused increased API error rates. The US-East-1 Region is based in northern Virginia.
The issue persisted through 3:30 pm ET (12:30 pm PT). At that time, AWS stated it was “pursuing multiple mitigation paths in parallel” and had seen “some signs of recovery” but could not provide an estimated time for full service restoration. At around 5 pm ET (2 pm PT), AWS said it had “executed a mitigation which is showing significant recovery in the US-EAST-1 Region” but still could not predict when service would fully come back online.
In a subsequent update, Amazon said API error rates and latencies recovered by 9:10 pm ET (6:10 PM PT) on December 7, with other lingering issues fully resolved in the early hours of Wednesday, December 8.
The outage hit a number of AWS services, including EC2, Connect, DynamoDB, Glue, Athena, Timestream and Chime, among others. In addition to impacting North America, the dashboard showed AWS Management Console and Support Center suffered increased error rates in South America, Europe, Africa, Asia Pacific and the Middle East.
Reports on DownDetector.com indicated the outage affected a number of popular consumer apps and services, including Netflix, Venmo, Disney+, game publisher Bethesda, Ticketmaster, IP-based phone service Mitel and remote education platform Canvas. Elsewhere, financial service firm UBS rescheduled several sessions of a virtual investor conference on Tuesday, citing issues with AWS which impacted its ability to broadcast the discussions live.
AWS suffered a similarly large outage in its US-East-1 Region in November 2020. CNBC reported the earlier event impacted Adobe, Roku, Amazon subsidiary Ring, lending company Affirm, Target deliver service Shipt, and Tribune Publishing.
The outage comes just after Amazon’s annual re:Invent conference, at which it touted a number of new products and services. This year it launched a private 5G product and touted work in the automotive, industrial and financial services segments.
This story has been updated to reflect that the issue which caused the outage has been resolved.