The outages started at around 6 a.m. ET, with users reporting they were receiving error messages including “Error 503 Service Unavailable” and “connection failure” when attempting to visit various websites.
The sites — which also included the U.K. government website — are now mostly back online.
Early reports have pointed the blame at Fastly, an American cloud computing services firm. Fastly’s status page said Tuesday morning that it was investigating a technical issue. It now says the issue has been resolved.
Shares of Fastly were down in U.S. premarket trading, but rose as much as 6 percent as markets opened for trading Tuesday.
What is Fastly?
Fastly operates what’s known as a content delivery network, or CDN.
CDNs are networks of servers and data centers distributed around the world that allow for the transfer of assets needed for loading internet content.
The company describes its technology as an “edge cloud” platform, which essentially means it places its infrastructure closer to the location where it’s needed to offer users faster response times.
Fastly’s tech is designed to counter common causes of online outages, such as distributed denial-of-service attacks which overwhelm a website with a sudden wave of traffic.
“The intention of CDNs is to route (or distribute) internet traffic and services through ‘nodes’ in order to balance the load of traffic, prevent bottlenecks and result in high availability and faster content delivery,” said Mark Henry, director of data protection and cyber security at global law firm DWF.
“Requests for content are directed by an algorithm, for instance the algorithm might direct the traffic so that it routes through the most available or highest performing node, or so that the traffic takes the fastest network route to the requestor.”
Henry said that some of the organizations affected by the outage were reverting to non-CDN distribution schemes to get their sites back up and running, but that this could result in a “slower than normal” experience for users.
Why did the outage happen?
The outage led to speculation online that Fastly may have fallen victim to a cyberattack. But Joshua Bixby, Fastly’s CEO, told the Wall Street Journal Tuesday that the outage wasn’t attack-related.
The company said in a tweet that it had “identified a service configuration that triggered disruptions” across its clusters of machines globally.
“Our global network is coming back online,” the company added.
Tuesday’s downtime was another reminder of the increased concentration of crucial internet infrastructure among a relatively small number of companies.
“This is what happens when half of the internet relies on Goliaths like Amazon, Google and Fastly for all of its servers and web services,” said Gaz Jones, technical director of digital agency Think3. “The entire internet has become dangerously geared on just a few players.”
When Amazon’s cloud computing unit Amazon Web Services encountered an issue in 2017, some of the world’s biggest websites went offline for several hours across the entire U.S. East Coast.
Meanwhile, in 2019 U.S. web security firm Cloudflare experienced a problem that lasted around an hour and affected websites including chat service Discord and the dating site OKCupid.