A large chunk of the internet was down, here’s what happened

If you felt like the internet was broken for a couple of hours on Thursday, you weren’t just imagining it. Many of the apps and websites you use every day suddenly stopped working. Services like Spotify, Discord, Snapchat, Character.AI, and even parts of Google itself seemed to vanish into thin air, leaving millions of users staring at error messages. The outage was highlighted by dozens of news outlets. Here’s a screenshot of Google’s news feed for reference:

So, what exactly caused this widespread digital blackout? The problem wasn’t a massive cyberattack or some shadowy international conflict, despite the online chatter. The real culprit was a bit more mundane, yet far more impactful. It was a failure deep within the internet’s plumbing.

The issue began with Google Cloud, one of the massive platforms that provides the foundational infrastructure for countless websites and online services. Think of it as a landlord for a huge chunk of the internet. When Google Cloud started having a bad day, it had a ripple effect across its many tenants.

One of its biggest tenants is a company called Cloudflare. You might not know their name, but they are a critical player in making the internet fast and safe. They handle an incredible amount of web traffic every second. As it turns out, several of Cloudflare’s own services depend heavily on Google Cloud’s infrastructure. So when Google’s platform stumbled, Cloudflare went down with it.

Cloudflare was quick to explain what happened. In a detailed breakdown, they revealed that a key service of theirs, known as Workers KV, experienced a massive failure. This service is essential for many of their other products, handling everything from configuration settings to user authentication. The storage infrastructure for Workers KV is partially hosted by a “third-party cloud provider,” which is Google. When that provider’s service failed, it set off a domino effect.

For about two hours and 28 minutes, the cascade of failures was significant. Cloudflare services that manage user access, security challenges, and even their own dashboard went dark. Because so many websites and applications rely on Cloudflare to function, they went dark too. The company noted that products like Access, which handles logins for many corporate tools, failed completely. Their WARP service, which helps secure user connections, couldn’t register new users. The list of affected services was long, touching everything from AI tools to video streaming. A recent report also highlighted indexing or serving issues with Google Search. So it’s clear the impact was widespread.

To their credit, Cloudflare didn’t try to hide or just blame Google. In a refreshingly honest blog post, the company took full responsibility. They stated, “We’re deeply sorry for this outage: this was a failure on our part.” They acknowledged that even though a third party’s issue triggered the event, “we are ultimately responsible for our chosen dependencies and how we choose to architect around them.”

On its status page, Google says it’ll publish an analysis of the incident once it wraps up the investigation into the matter. So I’ll post an update here once that drops.

Online, a flurry of outage reports hit sites like Downdetector. While users of Amazon Web Services also reported issues, the company stated its services were operating normally. The incident highlighted just how centralized the internet’s core infrastructure has become, with a few giant companies like Google, Amazon, and Microsoft underpinning most of what we do online.

By late afternoon, Google and Cloudflare managed to get things back online. The digital world returned to normal. Cloudflare says it has already started working on making its systems more resilient, hoping to avoid a single point of failure by not relying so heavily on one provider. I was fast asleep when all this happened, but let me know if you were impacted, and on which service in particular, in the comments below.

Featured image AI-generated