Major vulnerability exposed in Amazon’s global internet outage

It took a day without Amazon Web Services for Americans to realize how reliant the internet is on a single company.

It’s not just that people couldn’t place mobile orders for coffee at Starbucks or ask Alexa for the weather. Hospitals said crucial communications services weren’t working, and teachers couldn’t access their planned lessons for the day. Chime, a mobile banking service, was down, too, leaving people without access to their money. Ring and Blink cameras, along with most smart home devices, stopped working.

AWS is one of a small group of cloud computing juggernauts that form the backbone of the internet, providing businesses with backend computing tools needed to power crucial parts of their daily operations. That includes everything from storage to virtual servers that companies can use to develop and deploy apps without investing in their own hardware.

While other cloud providers exist, they lack the scale and reach of Amazon, Microsoft and Google. These three power the majority of the world’s cloud services, around 60% of the market, Roy Illsley, chief analyst at Omdia, told CNN. But among those three, AWS is the largest with roughly 37% of the market, according to research firm Gartner. AWS has a customer base of 4 million, according to an HG Insights report published this year.

So when a platform like AWS goes down, it has a cascading effect.

One expert already estimated the total impact of the disruption will be in the billions of dollars.

“It creates a very large single point of failure that then impacts operations at warehouses, deliveries, people being able to sell their goods and services to websites,” Jacob Bourne, an analyst at eMarketer, told CNN.

Hiccups at every step of the day

Debi Dougherty and her husband were affected by the AWS fallout on nearly every stop of their Monday morning errands in and around New Albany, Indiana.

To start the day, Dougherty was pinged with Ring alerts that there was a car in her driveway, but she couldn’t view the camera. She figured it was a Ring issue.

However, when at the doctor’s office for her husband’s first radiation therapy appointment, the scheduling software was so spotty that it took 40 minutes to book the next 25 days of appointments – something that usually only takes a few minutes.

The next stop at Kohl’s brought more delays. The line was backed up because the credit card reader was on the fritz.

The Doughertys then stopped for lunch at Cattleman’s Roadhouse, where the manager offered to pay for their meal because the restaurant was unable to process cards.

“He said, ‘This is no fault of yours, and you’re already eating. I don’t guess you all have cash?’” Debi Dougherty said. “And we both looked at each other, and I’m like, ‘Not enough to cover this meal.’”

Still, she said, the Monday morning experience was “frightening” to say the least, knowing how dependent society is on technology.

“(The businesses) put all their eggs in one (AWS) basket, because it’s affected so many different industries,” she said. “And, perhaps, that’s not the smartest thing to do.”

Rocky business

Cattleman’s, which uses the Toast point-of-sale system that’s reliant on AWS, was grateful that the outage happened on a Monday, and not a Friday, Saturday or Sunday, said Cameron Sharp, the New Albany location’s general manager.

“If this goes into multiday, or heaven forbid – and let me find some wood here so I can knock on it – goes into the weekend, we’re in trouble,” Sharp said.

Sharp ultimately had to comp just the one meal Monday before realizing that one Toast terminal at the restaurant could store the transactions.

“Our entire economy is based on e-commerce,” he said. “Because we’re so tied together, this (AWS outage) is going to screw with a lot of folks.”

Over in the broader Houston area, Dia Giordano was spending her Monday trying to untangle the mess that the outage made for her three businesses: an Italian restaurant, eight mental health clinics and a couple of rental properties.

DoorDash was “blowing up” her phone starting at 2 a.m., warning that the online ordering system, which is run through Toast, was down.

“What that means is one-third of my business is gone for the day,” she told CNN. “At least with the publicity (of the outage), people might be understanding, but I’m still getting messages asking if we’re open, because the website is just gone. It’s just not there.”

Toast, when reached Monday, declined comment.

At Giordano’s mental health clinics, her practitioners and administrative staff members were unable to validate clients’ insurance information because the online clearinghouse for that information wasn’t working.

And on top of that, Venmo was down, meaning she couldn’t receive the rental payments she normally would.