Cloudflare and AWS outages: The day the internet stood still

Email Security >

Cloudflare and AWS outages: The day the internet stood still

By Cian Fitzpatrick |  15th December 2025

Table of Contents

Introduction

The last quarter of Q4 delivered a shock to all of our systems with two recent outages. It’s no exaggeration to say that for a short stretch of time, it genuinely seemed as if the internet had slipped out of gear. 

In fact, what actually happened was far more specific. 

A small number of the infrastructure providers that quietly keep the online world humming ran into trouble. And because of how much the modern web leans on these companies, the ripple effect felt enormous.

Two outages equals one big wake up call

Amazon Web Services went down thanks to a bug in its automation software. Thousands of sites and applications were pulled into the chaos. 

Then, Cloudflare was the next to make headlines. The outage is estimated to have had a $1.6 billion impact on financial trading activity

This provider supports a huge amount of global web traffic. Inside Cloudflare, just one internal change was enough to knock a large portion of the internet sideways. 

The timing of these two incidents was uncomfortably close, and together the incidents highlighted something many people forget. The internet looks decentralised. However, its foundations are anything but.

What actually went wrong

Cloudflare’s issue began with what sounded like a routine configuration update in its bot management system. One file expanded far beyond its expected size. That growth triggered a failure in a piece of proxy software that directs traffic across Cloudflare’s global network. 

Once that slipped, the whole system wobbled. Requests across the network began returning 5xx errors in huge volumes. Websites that rely on Cloudflare’s CDN, security tools and edge platform quickly became sluggish or completely unreachable.

At one point, researchers estimated that close to one fifth of global web traffic was touched in some way.

AWS had a different problem. Yet, it had a similar overall effect. A fault inside its DynamoDB service in the heavily used US east 1 region triggered a series of DNS problems. Those issues then cascaded through other AWS services and caused widespread disruption.

If you depended on that region, you felt it immediately. More than a thousand online services experienced outages or slowdowns. Even some of Amazon’s own connected-home products were caught in the disruption. Many businesses could only sit and wait while AWS engineers brought the region back to stability.

The hidden fragility inside “the cloud”

These back to back incidents exposed a reality that rarely gets talked about. On paper, businesses believe they are protected because everything is in the cloud. But the cloud is not a single place. Nor is it automatically resilient. It is entirely possible to build modern systems that look sophisticated but rest on a fragile single stack.

Some of the most common patterns include:

  • Relying on one cloud provider, often in a single region, for most of a company’s production systems.

  • Routing almost all customer and API traffic through one content delivery or security provider, which means any failure at that layer takes everything with it.

  • Stacking multiple critical services on top of the same underlying infrastructure without realising it. A company could use tools that rely on AWS and are fronted by Cloudflare. Although they look like separate services, they’re actually one single point of failure.

When something deep inside this shared stack breaks, it breaks in a big way. The impact can be felt in logins not working, payments stalling and dashboards refusing  to load. Teams cannot even reach customers to explain what is happening. The fragility is only obvious once everything fails at the same time.

The cost of a standstill

Outages have always been expensive, but the stakes are even higher now. Companies now depend on online systems for nearly every part of their operations. Everything from sales to customer support to internal coordination relies on these digital foundations.

When core services go down, the consequences are big: .

  • Revenue slips away in minutes. E-commerce sites lose sales. Fintech transactions fail. SaaS companies face penalties or refund obligations. Emergency engineering work adds new costs.

  • Productivity drops sharply. When authentication systems, internal tools or communication platforms stop working, entire teams end up waiting for updates instead of doing their jobs.

  • Trust erodes. Customers rarely distinguish between a provider’s outage and the company they are trying to reach. If the service is down, it is the company that appears unreliable.

Reports tracking global internet performance now record more than one hundred major outages a year.

What resilience looks like today

Despite the recent experiences, Cloudflare and AWS remain among the most robust and well-engineered systems on earth. Both are critical to the infrastructure of our modern internet. 

However, modern resilience looks like this:

  • Diversifying across multiple regions or even multiple cloud providers. This ensures everything doesn’t sit in one basket.

  • Testing failover and recovery plans regularly. Don’t wait for a crisis to find out whether they work.

  • Preparing clear communication plans before you need them. Downtime is no fun for anyone, including your customers. Keeping them updated goes a long way to reassuring them that the problem is being solved.

Being ready both on the technical and operational front helps businesses stay steady no matter what disruptions come down the chute.

A turning point rather than a glitch

The Cloudflare and AWS outages were not freak occurrences. They were demonstrations of how intertwined our digital environment has become. A small change in one part of the web can now reach millions of users within seconds.

Contact our team today to discuss how to keep your business up and running no matter what happens.

Need help now?

Contact Us
Play sound

Join Our Topsec Newsletter Today

Sign up to get regular updates about email security

You have successfully subscribed to the newsletter

There was an error while trying to send your request. Please try again.

Topsec Cloud Solutions will use the information you provide on this form to be in touch with you and to provide updates and marketing.