Amazon’s cloud service, Amazon Web Services (AWS), resumed regular operations on Monday afternoon after an internet outage disrupted thousands of sites worldwide, affecting popular applications like Snapchat and Reddit. AWS, which provides application hosting and computing processes for businesses globally, suffered an interruption that impacted workers and halted regular activities such as online payments and ticket changes. Complaints of persistent difficulties with services like digital wallet Venmo and video-calling platform Zoom were reported on Monday afternoon.
Despite the resumption of services, Amazon noted that some AWS services had a backlog of messages that would require additional time to process. This isn’t the first time AWS has been implicated in a significant internet collapse. The northern Virginia cluster of AWS, known as US-EAST-1, has contributed to major internet meltdowns at least three times in the past five years.
Amazon did not provide a detailed explanation as to why this specific data centre continues to be affected. The recent problems were traced back to the Domain Name System (DNS), which averted applications from locating the correct address for AWS’s DynamoDB API, a cloud database essential for storing user information and other crucial data.
Earlier, AWS attributed the root cause of the outage to an underlying subsystem responsible for monitoring the health of its network load balancers, which help distribute traffic across multiple servers. The issue, according to AWS, originated within the EC2 internal network, Amazon’s Elastic Compute Cloud service, which offers on-demand cloud capacity within AWS. The issue was resolved around 3 pm PT (2200 GMT), although some services continued to have a backlog of messages to process.
Ken Birman, a computer science professor at Cornell University, emphasized the need for software developers to enhance fault tolerance, suggesting that AWS provides tools that developers can utilize to safeguard themselves in the event of an issue at one of its data centres.
As the world’s largest cloud provider, AWS offers computing power, data storage, and other digital services to companies, governments, and individuals. Disruptions to its servers can result in outages across websites and platforms that depend on its cloud infrastructure. According to AWS, Monday’s outage started at its US-EAST-1 location, AWS’s oldest and largest site for web services, which previously suffered outages in 2021 and 2020.
The problem underscores the interconnectivity of digital services and their reliance on a small number of global cloud providers. A single glitch can significantly disrupt businesses and everyday life.
The outage affected a vast range of companies across sectors. Apps like Reddit, Roblox, Snapchat, and Duolingo were all impacted. Other services such as Perplexity, a startup specializing in artificial intelligence, cryptocurrency exchange Coinbase, and trading app Robinhood also experienced disruptions attributed to AWS. Amazon’s own services, including its shopping website, Prime Video, and Alexa, were likewise affected.
What caused the AWS outage?
The outage was linked to an underlying subsystem that monitors the health of AWS’s network load balancers. It originated from within the EC2 internal network, Amazon’s Elastic Compute Cloud service.
What were the effects of the AWS outage?
The outage disrupted thousands of sites and applications globally, including popular apps like Snapchat and Reddit. It also halted regular activities such as online payments and ticket changes.
How often has the AWS northern Virginia cluster experienced major internet disruptions?
The northern Virginia cluster of AWS, known as US-EAST-1, has contributed to major internet meltdowns at least three times in the past five years.
