The internet just had another global outage. Why does this keep happening?

A significant disruption occurred on Monday due to an outage at Amazon Web Services (AWS), the cloud computing giant that underpins a large portion of online services. This incident rendered numerous major websites and applications inaccessible for several hours, affecting banking platforms, social media, airline bookings, and e-commerce sites. Millions of users, particularly on the US East Coast, found themselves unable to perform basic tasks like ordering coffee or using essential applications. This latest outage underscores the inherent fragility of the internet’s infrastructure, reminding users of just how dependent society has become on digital services. While AWS and its competitors typically boast robust systems, the intricate web of interconnected services means that reliability hinges on the performance of each individual component. The specific cause of the outage remains unclear, though it appears to be linked to issues with a service that translates web addresses into IP addresses, which hindered communication with AWS's extensive databases. Historically, outages of this magnitude can stem from various factors, including software updates, erroneous code injections, or compatibility issues with third-party software. In rare cases, physical disruptions like cable cuts or cyberattacks can also lead to such significant downtime. The frequency of these incidents raises concerns about the lack of necessary redundancies in the system, with some experts highlighting a tendency for companies to overly rely on a single cloud service provider. Rob Jardin, chief digital officer at cybersecurity firm NymVPN, confirmed there is no indication that this outage was the result of a cyberattack, suggesting it was likely a technical malfunction in one of Amazon's primary data centers. He noted, "The internet was initially designed for decentralization and resilience, yet a substantial part of our online ecosystem is now concentrated in a few cloud regions. When one of these regions fails, the consequences are immediate and widespread." Despite the rarity of such major disruptions, AWS has experienced similar incidents before, with the last significant one occurring in 2021. Cybersecurity expert Mike Chapple remarked that while disruptions like this attract considerable attention due to their impact, they are comparable in frequency to those experienced by other major cloud providers. In 2024, a severe IT outage caused by a software glitch in CrowdStrike led to widespread interruptions, including flight cancellations and disruptions in hospitals, resulting in substantial financial losses. Similarly, AT&T faced its own series of outages, including a significant 11-hour failure that hindered gig workers from fulfilling their tasks. AWS, which originated from Amazon's need for excess server capacity to manage holiday traffic, has grown into a cornerstone of online infrastructure, supporting many widely used services. The recent outage was largely tied to issues with the Domain Name System (DNS), a crucial component that helps users navigate the internet by turning user-friendly web addresses into comprehensible IP addresses. Although the data was securely stored, the inability to locate it temporarily severed connections between applications and their information, akin to a brief loss of memory for vast sections of the internet. The outage was resolved by 6:35 AM ET, with Amazon advising affected companies to clear their cache to expedite service restoration. As a precaution, AWS is expected to conduct a thorough review to determine the root cause of the DNS failure in the following days.