"We are investigating an increase in EC2 startup failures and network connectivity issues in one availability zone (USE1-AZ4) in the US-EAST-1 region," an Amazon Web Services (AWS) Service at 7:35 AM Eastern Time posted to the Health Dashboard.

"Other availability zones within the US-EAST-1 Region are not affected by this issue.

The issue was attributed to "power loss within a single data center." Power was restored by 8:39 a.m. Eastern time and affected services "should begin to recover." According to a 9:51 a.m. update, "the majority of AWS services have also been restored."

Online companies reporting service issues included Slack and Epic Games, and Down Detector showed users reporting problems with Grindr, Hulu, McDonald's, DoorDash, and others, although it was outages, it was not clear whether they were caused by the AWS outage.

According to Down Detector, some of these affected services appeared to have recovered by 9:30 AM ET. However, Slack reported five minutes later that there were "ongoing issues with file uploads," and at 10:06 AM Eastern time thanked users for "sticking with us while we continue to work on a fix."

According to Statista, AWS represents 32% of the total global cloud infrastructure market.

AWS had a similar problem on December 7, when one of its servers in the eastern U.S. went down, and again on December 15, when a server on the West Coast experienced problems. Today's glitch, as noted above, appears to be caused by another server in the eastern US.

In the two previous AWS outages, the problem seemed to be more widespread, affecting Disney Plus, Netflix, Sony's PlayStation Network, Amazon's own Prime Video, Alexa, Ring Video Doorbell service and even the navigation of delivery drivers was affected.

However, according to a Reddit thread, some of these services were affected this time as well.