- Jane Wakefield
- technology reporter
After a brief hiatus on Tuesday, major Internet sites including Amazon, Reddit and Twitch.
The British government’s service website has also stopped working, along with the websites of the Financial Times, the Guardian, the New York Times and others.
One message appeared on all of these sites, “Error 503 Service Unavailable”.
It seems that a glitch in the “Fastly” website, which provides online services to these websites, was the cause.
The company said the malfunction was caused by a problem with its global content delivery network (CDN), and it has begun to fix it.
“We identified an issue that triggered a malfunction in our Points of Online Presence (POP) on a global scale, causing the settings to be disabled,” it said in a statement.
Points of Presence allow content to be sent from globally distributed servers to the final destination.
Fastly works within what is known as the “Cloud” technology that supports increasing the speed of downloading files and sites, protecting them from hacker attacks, and supporting sites during peak browsing times.
It seems that the crisis has so far only included locations in the United States, Britain, and specific places in Europe.
Among the affected sites are PayPal, Shopify, the BBC, HBO Max and Vimeo.
Amazon and Cloudfire cyber services have had similar problems in the past.
And some sites were able to find solutions to the problem, as “The Verge” used the Google Docs service to spread its news. But the site forgot to specify which users are allowed to edit the content, which led to a series of funny edits and tweets.
The InternetOutage hashtag has spread widely on social media.
This has led some to question the rightness of concentrating so much of the corporate infrastructure on the Internet in the hands of a few companies, which means that one failure causes far greater crises.
“All of this highlights the importance, the danger, and the value of cyber-server companies,” said Jack Moore, a cybersecurity expert.
Adam Smith, a software expert with the BCS Institute, said the failure “sheds light on the complex system and the mating components that are involved in providing an Internet service, which is why outages are hitting multiple sites at the same time.”
“Outfalls like this are bound to happen from time to time, but they will be rare and short, and cloud providers build on the frequency of these events, to provide services that are more secure,” said Stephen Gilderdale, senior manager at Dell. “In most cases, services are affected only for a short time, and data can be recovered easily. Far from being a cause for concern, it shows the resilience of the network that can recover very quickly,” he added.