What really went down when the internet went down
Some people noticed the problem when they couldn’t access The Guardian. Others struggled with the New York Times or the UK government’s websites. Others couldn’t buy things Amazon. People started to panic as another global outage struck internet users.
The latest outage, which began just before 11am UK time, appears to have hit Fastly, a content distribution network, or CDN, and knocked out every company that used its services to support their websites. Across the internet, “Error 503 service unavailable” appeared on people’s screens.
Fastly identified the issue within 45 minutes and told the world that “a fix is being implemented”. The sites began to trickle back online soon after.
In less than an hour, a company that most people have never heard of showed how vulnerable our global internet infrastructure is. The vestiges of the internet not affected went into overdrive with speculation. While #cyberattack was trending Twitter, the reality was more prosaic: someone misconfigured a server. “We identified a service configuration that triggered disruptions across our POPs globally and have disabled that configuration, Fastly explained in a tweet. “Our global network is coming back online.”
“When you configure your servers, you give them commands. If you tell them something wrong, then it affects potentially all servers at once,” says Christian Kaufmann of Akamai, a competitor – who didn’t want to give specifics about Fastly. This isn’t the first time a configuration error has caused havoc. A 2020 outage at Cloudflare, another key CDN, occurred because of a configuration error, the company confirmed.
