Early last Tuesday morning, a large number of Internet users around the world were surprised by the message “An error occurred” and “Connection failed” when trying to access a number of Internet services.
The list included large sites and platforms such as “Amazon”, “Reddit” and “Target” and news sites such as “The New York Times”, “The Financial Times”, “CNN”, “The Guardian” and “Le Monde” and services such as “Pinterest” and HBO, Hulu, Spotify, and even the British government website, and the outage affected many countries in the Americas, Europe, Asia, as well as South Africa.
It turned out that the American cloud computing services company “Fastly” was responsible for this outage, which announced a “technical problem” that led to this failure, and later indicated that it fixed it and apologized to its users around the world.
What is “Fastly”?
The company operates what is known as a Content Delivery Network, or CDN. According to CNBC, CDNs are networks of servers and data centers distributed around the world that are designed to mitigate the traffic overload that can disrupt the service, as well as to counter any cyberattacks that favor disrupting it.
The company achieves this by storing content, websites and applications on servers close to users, according to CNN.
Technical expert Mark Henry told CNN: “CDNs aim to route or distribute Internet traffic in order to balance the traffic load, prevent bottlenecks, and lead to faster content delivery.”
But since Fastly supports a large number of companies and customers trying to access the various online platforms it serves, when the company is affected, users lose the ability to access these services entirely.
The outage led to speculation that it might have fallen victim to a similar cyber attack, but the company’s chief executive, Joshua Bixby, denied this had happened.
After being hit by a major pipeline and then the largest meat factory in the United States, it appears that the next target of the so-called “ransomware” hackers has become television stations.
What led to the malfunction?
As for the malfunction, it was caused by a software update that it did on May 12th. This update can cause a crash when a customer adjusts their service under certain circumstances, and the company said 85 percent of its networks experienced this issue.
After fixing the malfunction, which lasted about an hour, the company apologized and said it would work to prevent such widespread malfunctions in the future.
It confirmed that it will review its processes and practices to determine why the error was not detected, and will consider how to increase the speed of its network if another outage occurs.
Another problem reminder
The CNBC report says the outage is another reminder of the problem of concentrating so much of the Internet’s critical infrastructure in a relatively small number of companies.
Jazz Jones, Technical Director of Think3 Digital Agency: “This is what happens when half of the internet relies on companies like Amazon, Google and Fastly to use all their servers and web services…The entire internet has become dangerously oriented for a few players.”
Amazon’s cloud computing service had a similar problem, in 2017, that caused some of the world’s largest websites to stop working for several hours.
In 2019, US web security company Cloudflare faced a problem that lasted about an hour and also affected websites.