Large parts of the internet went dark on Tuesday morning after a widespread outage at Cloudflare left thousands of users locked out of their favorite apps and websites. The disruption hit heavy hitters like X (formerly Twitter), ChatGPT, Canva, and Grindr, leaving people unable to work or communicate for over an hour.
Cloudflare serves as the “plumbing” for a large portion of the internet, handling roughly 20% of global web traffic. Its job is to speed up site load times and protect sites from cyberattacks. Because they sit in the middle of so much data, when Cloudflare stumbles, the ripple effects are instant and massive.
The trouble started around 6:40 a.m. ET. Almost immediately, reports flooded in to Downdetector, a site that tracks web outages. Complaints peaked at nearly 5,000 as users found themselves staring at error messages instead of their feeds. Engineers at Cloudflare scrambled to identify the issue, eventually pinning the blame on a sudden spike in “unusual traffic” that overwhelmed one of their internal services “We are all hands on deck to make sure all traffic is served without errors,” the company explained in an email statement.
Fortunately for frustrated users, the team deployed a fix relatively quickly by 8 a.m. ET, the number of outage reports dropped to around 600, and services began returning to normal. However, the company warned that some customers might still see leftover glitches while the system fully recovers.
The financial markets reacted immediately to the instability. Cloudflare’s stock price dropped about 5% in premarket trading. This incident highlights a growing problem for the digital world: we rely heavily on just a few massive infrastructure providers. This crash comes just weeks after Amazon’s cloud service failed, which took down Reddit and Snapchat. As these outages become more frequent, they serve as a stark reminder of just how fragile the modern internet really is.
