Rapid Failure – Is the Internet Too Dependent on Cloud Hosting at the Edge?
- Cloud computing provider Fastly, which underlies many websites, said there have been issues with its global content delivery network (CDN)
- A huge amount of international news sites have crashed, including the New York Times, CNN and the BBC, while the stream of millions of pounds of income has been choked by companies such as Amazon, Boots and eBay.
Just minutes after an outage on the cloud platform Fastly provider on the Tuesday morning of June 9, and a significant portion of the Internet went extinct. People trying to visit a wide variety of websites received a blank white page and an error message telling them that the connection was not available. The blackout underscores how important little-known internet infrastructure companies like Fastly are, and how even isolated cloud disruptions at the edge can shut down large aspects of online business that we typically take for granted.
the San Francisco-based cloud service provider disruption was relatively brief, lasting a little over an hour in most cases. Fastly later said on his website that he had identified a problem and applied a fix, allowing the sites concerned to come back online. But the brief hiatus impacted a collection of top-notch web portals so quickly, including well-known news sites and even government portals like the UK’s “.gov.uk” platforms, that it recalled other disruptions in edge cloud hosting. services like Amazon Web Services (AWS) and Cloudflare – highlighting just how vulnerable the underlying web infrastructure can sometimes be, even for the largest operators.
What is Fastly and what caused the internet to go down?
The company offers a content delivery network or CDN service. When it works, a CDN is supposed to improve the speed and reliability of the Internet. Rather than visitors to a website all having to log into servers run by that company – which may not even be in the same country they are – they instead contact Fastly, which runs huge server farms in all over the world who host copies of their customers’ websites. .
This means that although hosted in the cloud, the page loads faster for the user because it is located closer to those edge server farms, the physical signals don’t have to travel that far. It also improves website reliability, ensuring that if there is a big spike in traffic, it hits Fastly’s servers first, which are designed to handle heavy traffic loads.
We have identified a service configuration that has triggered disruption to our POPs around the world and have disabled that configuration. Our global network is coming back online. Continuous status is available on https://t.co/RIQWX0LWwl
– Quickly (@fastly) June 8, 2021
As to the cause of the internet outage, the problem was quickly identified and within minutes, Fastly had admitted on a status page that he was having problems. With the exception of a few suppliers, including the BBC which had backup systems, each affected website had to wait for Fastly to correct the error before they could restore the service.
Fastly is one of the few major CDN providers: the others include Amazon’s Cloudflare and CloudFront. But, to give an idea of Fastly’s reputation in cloud web hosting circles, Amazon’s retail website actually goes through Fastly, rather than CloudFront, and has been since May 2020.
Nick Rockwell, Senior Vice President of Engineering and Infrastructure at Fastly in a blog post said, “We experienced a global outage due to an undiscovered software bug that emerged on June 8 when it was triggered by a valid client configuration change. We detected the disturbance within a minute, then identified and isolated the cause and disabled the configuration. Within 49 minutes, 95% of our network was functioning normally. “
Who was impacted by this?
Among the large organizations affected by Tuesday’s internet blackout were the New York Times, Amazon, and Hulu. Other news outlets affected by the blackout included CNN, the Guardian, Bloomberg News, the Financial Times and The Verge. Apparently all UK government websites using the ‘gov.uk’ suffix have also been taken down, restricting access to public services.
Another well-known service to suffer was Spotify, the music streaming platform. The Reddit discussion board site has seen disruption, as has the image-sharing site Pinterest, live game streaming site Twitch, and video-on-demand platforms, Hulu and HBO Max. In some cases, the issue affected some services on platforms that were otherwise unharmed. Twitter users were briefly unable to use the emojis because the servers that host them were affected.
This is not the first time that a cloud service provider drew attention to the dangers of putting all your eggs in someone else’s basket. Earlier this year, AWS unilaterally suspended the service of the social media site Speak for largely arbitrary reasons that seemed to have more to do with political tribalism than anything else. Even then, the British government was unable to digitally serve its own people for some time.