What Triggers Global Web Outages and How Businesses Can Evade Them

For most modern-day companies, business operations often pivot around their websites and apps. They allow them to connect with customers, sell products and services, and gain new insights that can be used to improve customer experience (CX). Given the importance of online channels, web outages — even for a few minutes can be bad for businesses. We look at the recent website outages, what triggered them and how companies can mitigate their impact. 

On June 17, several major Australian banks, including ANZ, Allianz, Commonwealth Bank, and some leading U.S. airline companies such as American Airlines, Southwest Airlines, United Airlines, suffered a brief outage. Customers couldn’t sign into their accounts while the Reserve Bank of Australia had to cancel plans to buy long-dated government bonds. 

This was the second major global outage involving multiple websites in ten days. On June 8, websites of e-commerce companies Amazon and eBay; social media networks Twitch and Reddit; music streaming company Spotify; news media companies such as Financial Times, the Guardian and the New York Times; and the website of the U.K. government also crashed and remained inaccessible across the globe for a brief period.

Website outages have become more frequent than ever. According to DowndetectorOpens a new window , on average, more than 50 online services, including leading websites, online stores and streaming platforms, suffer an outage almost every month. 

Learn more: 3 Ways Retailers Can Leverage the Cloud To Provide Rich Brand Experiences  

What Causes Website Outages?

1. Issues in the content delivery network

Many of the recent website outages, including those reported in June, were triggered by technical issues in content delivery networks (CDNs). 

For the June 8 outage, UK-based Fastly said it was caused by an undiscovered software bug that surfaced after a valid customer configuration change. The June 17 outage occurred due to an inadvertent increase in the routing table value used by Prolexic DDoS service, resulting in unanticipated service disruption, stated CDN provider Akamai.

CDNs have become a critical component of modern-day Internet infrastructure. They reduce a website’s loading time by keeping a copy of in-demand content on a highly optimized network of servers spread across the world. 

“CDNs generate replicas of original websites for the website owners to allow load balancing. So instead of everyone worldwide accessing one centralized server and causing an overload, what they do is spread the load between different replicas,” explains Lotem Finkelstenn, Head of Threat Intelligence at Check Point Software Technologies.

“Everyone is routed to the nearest server to their device, and when a CDN fails, it means that all the replicas are unavailable and no one can see the content from the original server,” adds Finkelstenn. 

Finkelstenn has a point. A handful of CDN companies now support many of the world’s most visited websites. So if they go down, many online services dependent on them go down too at the same time. 

2. Cyberattacks

Large-scale cyberattacks can also cause website outages. A case in point is the notorious Mirai botnet attack of 2016. Attackers used millions of hacked IoT devices to carry out a massive distributed denial of service (DDoS) attack on systems operated by Dyn, a primary Domain Name System (DNS) provider. In what is now considered one of the biggest cyber attacks, a major part of the Internet on the east coast of the U.S. went down for hours in a blink. Close to 100 leading websites suffered an outage across North America and Europe. Twitter, Netflix, Amazon, Airbnb, Reddit, Slack, Verizon Communications, GitHub, CNN were some of the companies affected during the outage. 

Learn more: How SMBs Can Stop Damaging Cyberattacks and Reduce Downtime 

3. Sudden surge in traffic

A sudden unanticipated increase in traffic can also trigger website outages. It typically occurs due to a mismatch between the traffic levels at any point in time and the capacity of the website’s infrastructure to handle it. A case in point is the release of the video game Cyberpunk 77 in 2020, which led to a large volume of traffic on Steam and brought down the gaming store’s website on December 9.  In the U.S. alone, there were 29,212 issues reported on Downdetector.Opens a new window

4. Issues with primary cloud providers

In 2017, online services of several companies, including Quora, Trello, IFTTT, and websites created with Wix, suffered an outage. They remained inaccessible for several hours after some of Amazon’s S3 web servers went down. Amazon later revealed that on the day of the outage, the company was debugging the billing system, and for that, they had to turn off a small number of servers. During the process, one team member entered an incorrect command, resulting in the erroneous removal of a more extensive set of servers. A full restart was initiated, and it took a while before the services were restored. Glitches like these are rare, but they have led to some of the outages in the past. 

In November 2020, AWS again suffered a global outage that lasted several hours and interrupted services for Adobe Spark, Autodesk, Flickr, iRobot, Roku, The Washington Post, and Glassdoor. 

How Website Outages Can Impact Companies?

More than 85% of Americans have access to the Internet as of 2020, according to StatistaOpens a new window . The growing Internet penetration has made apps, websites and web services critical tools for communication, lead generation, sales and productivity for most businesses. When Google services went down for 6 hours in August 2020, several key productivity apps such as Gmail, Meet, Docs, Sheets and Drive stopped working. Businesses across the world rely on these services for day-to-day communication and operations. 

Likewise, when Amazon’s e-commerce store went down for nearly an hour in 2018 during one of their Prime Day Sales, the company lost close to USD 100 million, according to an estimate by retail discount aggregator Lovethesales.com, which was tracking prices and discounts on the e-commerce store during prime day sales. 

According to Uptrends, an uptime monitoring platform, a website’s uptime already drops below 99.99% if an outage lasts an hour. If it takes 3-4 hours to fix, it slips further to 99.95%. When the outages become frequent, the uptime can drop to 90% or lower, which can be bad for business. 

“It is hard to know the exact cost of these outages, but given the size of some of the web properties that were affected, we can safely assume it wasn’t zero,” said Doug Madory, Director of Internet Analysis at Kentik, a network observability company. 

Madory points out, the consolidation of the Internet industry and dependence on a few cloud providers have widened the impact of these outages. 

Learn more:  6 Cloud Data Management Hacks to Optimize Both Cost and Performance 

How Businesses Can Avoid Website Outages

Website monitoring: Deploying a website performance monitoring tool such as Pingdom, Uptime Robot at their end can come in handy. These tools record downtime, uptime and response time and also verify downtime from multiple locales. 

Backup DNS

Using backup DNS servers allows website owners to keep their website accessible even if the primary servers go down. In a blog postOpens a new window , John Bowers a J.D. candidate at Yale Law School, and Jonathan Zittrain a Professor of Computer Science at the Harvard School of Engineering and Applied Sciences, wrote, “Despite a slight uptick in the practice following the Dyn attack, the use of secondary DNS—which could enable websites to diversify their DNS across multiple providers, or self-host backup servers—remains marginal.”

“It has the potential to substantially increase resilience to error and attack on an internet-wide basis. External DNS providers should encourage the use of secondary DNS, making it as easy as possible to configure a backup provider,” they added.  

Multiple CDNs

 Another trick that can come in handy during downtimes is using more than one CDN provider. “There is little that an individual website operator can do to prevent a CDN from suffering an outage. One option is to use multiple CDNs so that traffic could be shifted if a CDN were to suffer an outage,” pointed out Madory.

However, signing up for multiple CDN services can be costly and may appeal only to large websites with deep pockets.  “It may not be cost-effective to operate two CDNs if the outages are sporadic and short-lived. It depends on the sensitivity of the business – while a large e-commerce site may justify this level of resilience, a smaller site may need to live with rare outages,” he added.  

Conclusion

The complexity of modern-day internet infrastructure has made website outages an unavoidable reality. There is no full-time solution in sight now, but businesses can mitigate their impact with some policy changes. 

What are your thoughts on the frequent global web outages and the disruption they cause. Comment below or tell us on LinkedInOpens a new window , TwitterOpens a new window , or FacebookOpens a new window . We would love to hear from you!