Snapchat, Signal, Duolingo, Roblox, Coinbase, Fortnite, Epic Games. They are just the most popular of the hundreds of platforms that went offline around the world after a problem with the Amazon cloud, the largest provider in the world, on which they rely.
“We can confirm significant error rates,” explained the company, which has been working to gradually mitigate the problem. It’s just the latest example of the fragility of the global internet infrastructure. where a failure to one widely used piece of technology can bring systems to a halt around the world. The global outage began after 9am Italian time and also affected our country.
Many users have complained about the downtime of various online services, also documented by the Downdetector site which monitors online faults: according to the BBC it has received over 6.5 million reports and more than 1,000 companies have been affected.
These include Perplexity which deals with AI and the gaming platforms Roblox, PokemonGo, Clash Royale, Clash of Clans and Brawl Stars. Apps like Canva for graphic design; Snapchat and Signal chats; the social network Reddit and various video streaming platforms such as Hulu and Disney+, the cryptocurrency exchange site Coinbase. In Europe, several mobile operators and some UK banks were also affected.
Services linked to Amazon were also affected by the malfunction: from the popular e-commerce site to the Alexa voice assistant. And, above all, Amazon Web Services (Aws): it is the cloud service of the technological giant, the infrastructure underlying millions of large company platforms. Many of the apps on our smartphones are actually running on Amazon Web Services data centers, which are increasingly overloaded also due to artificial intelligence processing.
“We can confirm significant error rates for requests directed to the DynamoDB database,” AWS said in an update on the service page, adding that the problems were concentrated in one of its “endpoints for services in the US-East-1 region,” which is located in Virginia. “Most of the operations of the AWS services are now performed normally”, he then specified, referring to “the mitigation of the DNS problem”.
It is the Domain Name System, often compared to an Internet telephone book: in practice it translates the names of the sites used by users into the numerical equivalents of IP addresses that can be read and understood by computers. Its interruptions prevent online browsing programs, browsers, from locating the content they are looking for.
The outage showed “how dependent we are on companies like Amazon, Microsoft and Alphabet for many of the online services that we more or less take for granted,” analyst Michael Hewson told the Guardian, “economically it’s almost like putting all your eggs in one basket.” Among the most sensational recent outages, there is the case of CloudStrike in July 2024: an update error in the IT security software sent Windows computers around the world into a tailspin. While in October 2021 a configuration error blocked Facebook, Instagram and WhatsApp for almost six hours.
WHAT IS AWS
Born in 2006 as an internal Amazon project, AWS is today the world’s leading cloud computing provider: it manages servers, databases, artificial intelligence systems, data storage solutions and virtual networks for companies, public bodies and startups. In total we are talking about over 200 services which progressively increase. In essence, Amazon rents computing power and digital space “on demand”, allowing customers to pay only for actual use, without having to physically own or maintain the servers.
The advantage is clear: instead of purchasing and operating expensive infrastructure around the clock, companies can “rent” digital resources such as computing capacity, data storage or databases directly via the Internet, paying exclusively for what they use. A model that has made Jeff Bezos’ company the world leader in the sector, with over 30% of the global cloud market share. According to data from Synergy Research Group, Azure (20%), Google Cloud (13%), Alibaba Cloud (4%) and Oracle (3%) follow.
WHERE THE AWS SERVER AREAS ARE LOCATED
AWS operates in over 30 regions around the world. The one in Virginia, at the center of today’s malfunction, is one of the largest and most strategic. The global network includes operational areas in North America, Asia and the Middle East – including Mumbai, Osaka, Israel and the United Arab Emirates – and of course in Europe, where regions in Ireland, Germany, France, Spain, Sweden and also in Italy, with the Milan region, are active.
THE CUSTOMERS
Amazon Web Services’ major customers include some of the most influential companies and institutions on the planet. Here are some examples: Netflix uses AWS to distribute content to millions of subscribers around the world with minimal latency, ensuring streaming continuity even during traffic peaks. NASA, through its Jet Propulsion Laboratory, uses AWS to process enormous quantities of data from space missions. Capital One, one of the main US banks, has also migrated its systems to the Amazon cloud to offer “cloud banking” services with high standards of security and regulatory compliance. Added to the list are companies such as Siemens, Lloyds Bank, BMW, Pfizer, AirBnb, Adobe and Spotify, but also Slack, Snapchat and Reddit among the social networks, and then Fortnite, Uber and Perplexity. All companies that use AWS to enhance data analysis, artificial intelligence and new generation digital services.
THE RISKS
The centralization created by structures like AWS can, as seen today, pose systemic risks. When an AWS “region” crashes, due to cyber attacks or malfunctions (and it is still early to understand what really happened) a domino effect can affect thousands of services. Finally, today’s blackout confirms how dependence on the cloud is now structural for the digital economy and how the security of digital infrastructures has become a major geopolitical and industrial issue