For a company pitching itself as helping improve the reliability and speed of websites, it is hard to imagine a worse turn of events than what Fastly experienced this morning.
Fastly, a cloud-computing company used by businesses around the world to operate their websites, faced an outage that caused popular websites including Reddit, Twitch, Hulu and The New York Times to suddenly crash for about an hour early today. Websites in North and South America, Europe, Asia and Africa were affected, the company said.
Bay City News, a Bay Area online news service, reported that Fastly acknowledged on its website at 2:58am. Tuesday that it was “currently investigating potential impact to performance with our CDN services.”
Numerous sites, including some San Francisco news sites, were reported down, including The New York Times, CNN, Spotify, BBC and some UK government websites.
By 3:44am, Fastly reported that it had identified the issue and a fix was being implemented. Bay City News reported that the SFGate and the San Francisco Chronicle websites were observed down at 3:40am with error messages displayed instead of news content, but both sites were back up at 3:46am.
Fastly posted another update on its website at 3:57am.: “The issue has been identified and a fix has been applied. Customers may experience increased origin load as global services return.”
Fastly, based in San Francisco, provides the kind of behind-the-scenes technology that most people do not know exists, but is crucial for making the internet work. The company had been enjoying some success. Its stock skyrocketed last year, benefiting from all the people who were online amid the pandemic, the same wave that helped other tech companies like Zoom before coming back down over the past several months.
Before the system crashed today, June 8, the Fastly stock had more than doubled since the company went public in 2019. In premarket trading, its stock briefly tumbled more than 7 percent before settling at about 1.6 percent lower.
Fastly provides a technology known as a Content Delivery Network, or CDN, a highly distributed network used to reduce the distance between a server and a user, thus accelerating website loading speeds. Fastly says its network improves reliability because it distributes a website to many locations, rather than depending on a central data center.
Everything began going sideways for Fastly on Tuesday morning. At 5:58 a.m. Eastern time, the company posted on its website that it was investigating a problem with its services.
Customers that were experiencing outages included PayPal, The New York Times, The Financial Times and The Guardian. Even the British government’s main webpage was offline.
A spokeswoman for Fastly, whose own website was taken down, attributed the problem to a disruption to Fastly’s systems that send website content from a server to individual users.
Kentik, a network analytics company, said internet traffic volume coming from Fastly’s services fell by 75 percent.
The complexity of the internet means that short-lived outages will always exist, said Doug Madory, Kentik’s director of internet analysis. He noted that Google, Amazon and other large companies have also experienced problems that take services offline.
“There is no error-free internet, it doesn’t exist,” he said. “This is unavoidable because of the complexity.”
This article originally appeared in The New York Times, copyright 2021. Bay City News contributed to this San Jose Inside report.