Services including Apple Music were among the hundreds of websites and apps that were down spanning the entire internet because of an Amazon Web Services problem.
This isn't the first time Amazon Web Services (AWS) has gone down, but it is another example of how dependent the entire internet is on this one service. As of 5:30 A.M. Eastern, Amazon says on its status page that sites should be returning to normal, but it isn't clear how long it will take before all apps and sites are back.
Apple's status page, however, claims that all problems have been resolved. It specifically says that Apple School Manager and iCloud Storage Upgrades had problems during this time.
Users of Downdetector, though, reported problems with Apple Music, Apple TV, Apple Card, Apple Pay, and more.
Apple is a significant user of AWS. As of 2019, it was reportedly spending over $30 million per month on AWS. That's despite recurring suggestions that it might build its own rival.
What happened
The issue began at around 3 A.M. Eastern and appears to have centered on Amazon's North Virginia AWS site. Amazon now says that the problem concerned DNS errors and was confined to what it calls its US-EAST-1 region.
However, services across the US and also Europe were affected, with the UK's government site, the EE carrier, and Zoom being offline. Some services such as Lloyds Bank, posted notices saying certain features were down, while Duolingo claimed it was undergoing maintenance.
As of 6 A.M. Eastern, sites and services including all of Apple's, plus the Epic Games Store, and Amazon itself, appear to be functioning normally. Others including Duolingo and certain banks are yet to be fully restored.
Beyond saying that the issue was related to the "DNS resolution of the DynamoDB API endpoint in US-EAST-1," Amazon has not yet reported how the problem began. When the same US-East-1 region went down in 2017, it was eventually found to be because of human error.
