The Internet is now an essential household tool for many Americans, perhaps even on a par with running water.
It turns out that the pipes of the internet can also leak.
In the past three weeks, two major disruptions to Amazon’s cloud computing services have caused widespread disruptions to other online services. Last month, a problem at Comcast, one of the largest Internet service providers in the US, led to widespread outages. (Comcast owns NBC News.) And in June, websites around the world were temporarily taken offline when Fastly, a cloud computing service provider, “service configuration” issues addressed†
The drumbeat of trouble underscores that despite all it’s capable of, the Internet is vulnerable at times.
“It’s expected to be like your power or your water, and sometimes they go out,” said Steve Moore, the chief security strategist at cybersecurity firm Exabeam.
The latest outage occurred on Wednesday, when customers of DoorDash, Hulu and other websites complained that they couldn’t connect† The problems were attributable to Amazon Web Services, or AWS, the most widely used cloud services company, which reported outages in two of its 26 geographic regions impacting services nationwide.
A similar outage occurred on December 7, crippling video streams and shutting down internet connectivity robot vacuum cleaners and even shut down pet food dispensers in a series of reminders of how much life has gone online, especially during the coronavirus pandemic. AWS published a unusually detailed description of what went wrong, along with an apology.
The incidents helped shatter the illusion, reinforced by decades of steadily improving internet speed and reliability, that everyday consumers can rely on online services to be available without fail.
In the old days, online video meant watching “a low-resolution video for five minutes,” said Robert Blumofe, executive vice president and chief technology officer at Akamai Technologies. Akamai sells security services and edge computing capabilities, a type of distributed technology that doesn’t rely so much on centralized data centers.
“Now there’s a very strong expectation that you can watch an entire movie in high resolution,” Blumofe said. “There is a recency bias. We remember the immediate and the now more than we remember how things were in the past’, when there were frequent disturbances.
In other words, some Americans who enjoy reliable Internet access may have gotten a little spoiled.
Computer science and security experts said the disruptions don’t really question the fundamental design of the Internet. One of the rationales was that a distributed system usually continue to function even if a piece goes down.
But they said the problems are rooted in the uneven development of the Internet, as certain data centers are more important than others; cloud companies from Amazon, Google and Microsoft concentrate more power; and cloud services business customers don’t always want to pay extra for backup systems and employees.
Sean O’Brien, a cybersecurity lecturer at Yale Law School, said the power outages cast doubt on the wisdom of relying so much on large data centers.
“‘The cloud’ has never been sustainable and is just a euphemism for concentrated network resources managed by a centralized entity,” he said, adding that alternatives such as peer-to-peer technology and edge computing could come into favor. He wrote after last week’s outage that the major cloud providers are “feudal system†
Cloud service providers make money by selling server space to other companies on flexible terms and with specialized expertise, reducing the need for companies to manage their own servers. They rarely fail, but they get attention when they do. An AWS outage in November 2020 affected customers like Apple†
“There are many points of failure whose unavailability or sub-optimal operation would affect the entire global experience of the Internet,” said Vahid Behzadan, an assistant professor of computer science at the University of New Haven.
Some of those failings — like AWS’s “us-east-1” region — have become notorious among tech workers who share their experiences on industry bulletin boards†
“The fact that we’ve had repeated outages in a short period of time is cause for concern,” said Behzadan, noting that US companies placed a high value on the assumption that cloud services are resilient.
He also said that as outages become more frequent or publicly visible, enterprise customers are likely to respond by spending more on backup systems to ensure they are resilient in the event of outages — with contracts with both Google and Amazon, for example. There is now a new industry debate on whether or not to go “multicloud”, CNBC reported:and companies in different sectors spend more on edge computing tools†
“The internet will not disappear anytime soon. But what doesn’t kill the internet makes it stronger,” Behzadan said.
Moore, of Exabeam, said the tightening job market across the country could also impact cloud services and internet reliability, as any increase in customer churn reduces the experience level of the people in charge.
“We are emerging from unprecedented times when people are incredibly stressed and expectations for cloud infrastructure are higher than ever,” he said. “Organizations are catching up.”