Copyright WJLA
A took many major websites and apps offline for several hours early Monday morning in an outage that highlights the sensitivity of the online infrastructure that powers much of the world’s technology. Amazon said the outage started shortly after 3 a.m. Eastern and had been mostly resolved at 5:27 a.m. Most services and sites had been restored as the East Coast hit regular business hours, but intermittent issues lingered throughout the morning and into the afternoon with additional “connectivity issues” in the US-EAST-1 region based in northern Virginia. A notice on the Amazon Web Services status page said it was experiencing issues with DynamoDB, a database that hosts information for companies. Its clients couldn’t access the data stored there because the Domain Name System — which translates user-friendly web addresses into IP addresses so sites and apps can load — encountered a problem. Amazon did not say what caused the DNS outage, which was resolved a few hours after the original issue. Cybersecurity experts said there were no signs that the disruption was caused by a cyberattack. The disruption impacted dozens of major services and companies like WhatsApp, Venmo, Hulu, Snapchat, and McDonald’s that also captured government services abroad with the British government’s website and tax services facing outages. It also effected consumer products like Ring doorbells and smart household appliances that need internet connection to fully function, adding another headache for customers to deal with during the disruption to service. “Many of the applications and services we depend on in our daily lives, including email, messaging, social media, banking, online shopping, travel booking, music and video streaming, etc. are all implemented as web services,” said Tevfik Kosar, a professor of computer science and engineering at the University at Buffalo. “When any of these cloud service providers face any issue in their systems, this impacts hundreds of millions of people, just like the AWS outage we experienced today.” Down Detector, a site that tracks internet outages, said there had been more than 8 million reports connected to the AWS outage by mid-morning on Monday. AWS is one of the largest providers of cloud-based services with sprawling infrastructure to support it around the world with thousands of clients that rely on it for complex and data-intensive operations like streaming video, storing data and running online applications. Amazon, Microsoft and Google provide most of the support for how much of the internet functions, meaning even minor issues for one company can quickly become more widespread and effect millions of people. Monday’s outage was the latest high-profile reminder of that issue. Last year, by little-known cybersecurity firm CrowdStrike forcing airlines to ground flights, hospitals to cancel surgeries and beyond with a daylong outage prompted by a small strand of faulty code. AWS has also had outages before in 2023 and 2021 that led to issues with people being able to access things like airline reservations and payment apps. Companies have increasingly turned to major players like Amazon and Google to handle their needs instead of having their own data centers, which was more common in the past. Using outside services instead of their own allows companies to adjust scale of operations without costly investments in hardware but comes with the trade-off of relying on them for functionality of their businesses. Amazon, Google and Microsoft’s services are typically very reliable and are quickly fixed, meaning disruptions are usually limited to a matter of hours rather than days. But experts said the latest issue highlights the need to build more resilience into the complex system that upholds most of modern business and society. “Any application or service which solely depends on a single cloud service provider is prone to “single-point of failure syndrome”. Its accessibility, availability and reliability can suffer anytime. This can happen sometimes due to very simple technical errors at the cloud service provider, sometimes due to local power outages, and sometimes due to bigger problems, such as regional blackouts or cyberattacks,” Kosar said. Some companies have spread their online services between multiple cloud providers to create a backstop in case of an outage or other technical issues, but Monday’s AWS outage put a spotlight on how many companies are relying on third parties for their infrastructure. “So many online services rely upon third parties for their physical infrastructure, and this shows that problems can occur in even the largest of those third-party providers. Small errors, often human made, can have widespread and significant impact,” said Konstantinos Mersinas an associate professor at the University of London.
