Internet outages: Why you can expect more to come
With so much of our lives digital, even a glitch like one Tuesday can be devastating.
The internet took a hit early Tuesday with a short-lived but massive outage shuttering online destinations as diverse as Amazon, Google, Reddit, Twitch, USA TODAY, CNN, The New York Times, the Financial Times and the U.K. government website.
Within an hour or so, cloud content company Fastly corrected the error and sites began operating again.
Expect such outages to be the norm because this wasn’t the first time – nor will it be the last – in which online outages wreak havoc with our lives.
With more and more data and services moving online amid a growing network of computer hubs across the U.S. and the world, issues will arise due to glitches and mechanical failures – or worse, from bad actors such as hackers and ransomware purveyors.
Still, we continue to adopt an increasingly digital lifestyle, with more functionality on mobile devices – Apple on Monday proposed putting driver’s licenses, as well as home and car keys, into iPhones. And most of us don’t really think about or understand, the technology behind this digital lifestyle.
Plus, this connected existence is not as robust, reliable and secure as you
might think. Just as subways may run slower than expected or trains derail, so too can there be incidents on the information highway.
“It’s a scary reminder of the doubleedged sword around the digital transformation,” said Daniel Ives, managing director of equity research at Wedbush Securities. “It’s just a few dominoes that could shut everything down.”
Over the years, we have grown to expect Netflix and Disney+ to almost instantaneously deliver “The Queen’s Gambit” or “The Mandalorian” with a click. And we expect news to be available on demand 24/7 on sites such as CNN, Bloomberg, or CNET – destinations also affected Tuesday.
“We just assume all this stuff is here all the time. I think the purveyors of our digital lives have gone out of their way to made us feel like it’s always there,” said Shelly Palmer, CEO at The Palmer Group, a tech strategy advisory group, and author of “Blockchain – Cryptocurrency, NFTs & Smart Contracts: An executive guide to the world of decentralized finance.”
Amazon CEO Jeff Bezos “has reduced every ounce of friction from you buying something. Mark Zuckerberg has reduced every ounce of friction they can about you posting something on a social network,” Palmer said. “The only time people think about this experience is when it goes away.”
How do shutdowns happen?
This latest incident occurred due to a problem at Fastly, a San Franciscoheadquartered content delivery network that supports websites for companies such as The New York Times, GitHub, Pinterest and others. It has deployed thousands of computer servers across the U.S. and the globe so that content is easily accessed.
There are just a few major content delivery networks such as Fastly, including Amazon, Akamai and Cloudflare. Each has redundancies built into their networks, but problems can arise. An Amazon Web Services outage in November 2020 took down the video game “League of Legends” and Sirius XM satellite radio and also affected Roku and Amazon’s Ring doorbell. AWS had similar outages in 2015 and 2017.
“I think that these crash so infrequently, it’s news when it happens,” Palmer said. “The goal here is speed … You want to see your video immediately pressing a button. You want everything to work beautifully and smoothly. The way you do that is you cache (or store) content as close to the user as possible. That’s what a content distribution network does.”
Normally, the work a content delivery network does, helps “content pop up on your screen in seconds,” tweeted Corinne Cath-Speth, a doctoral student at the Oxford Internet Institute, in a description of how such outages happen.
“So when those @fastly services fail or falter – it has major ramifications for everyone’s internet experience. Keep in mind that this market is relatively small and each company serves a large number of clients,” she tweeted.
All this works as it should more than 99% of the time. How much more would a company have to spend to improve that to nearly 100%? Probably too much, Palmer said.
What can you do?
Each of us should use this recent event to consider our situation. Think about how often there are online outages that could affect you. Connectivity programs such as Microsoft Teams and Slack have had outages recently. So have social networks such as Facebook and Instagram. Do you have phone numbers or emails for co-workers, friends, or family you might need?
Many of us store personal files in the cloud and those networks such as Google Cloud and Apple’s iCloud can have outages, too. You might want to have multiple ways to save important files, photos and other data. In addition to storing them in the cloud, have them on an external drive or USB drive.
If you have more than one computer, have copies on both devices in case one is infected. And consider encrypting files for added protection.
That could come in handy if you become a victim of ransomware or malware, as fuel supplier Colonial Pipeline and meat producer JBS S.A. did. Said Palmer: “These are teachable moments about how vulnerable we are … and how deeply we have come to rely on our connectivity and how out of control of it we really are.”