A wave of major websites — including LinkedIn, Zoom, Canva and Shopify — briefly crashed on Friday morning after Cloudflare suffered yet another outage, marking the company’s second disruption in under a month.
Cloudflare said the issue was unrelated to any cyberattack. Instead, it was triggered after the company adjusted its firewall to defend customers from a newly discovered software vulnerability. The company also acknowledged an earlier problem with its application programming interfaces.
The disruption lasted around 30 minutes and was resolved shortly after 9 a.m. GMT. It comes just weeks after a much larger outage in mid-November, which brought down X, Spotify, OpenAI, and several multiplayer games like League of Legends.
That earlier incident was caused by an oversized configuration file that unexpectedly crashed systems responsible for handling Cloudflare’s traffic.
Friday’s glitch was far smaller but still knocked several high-profile sites offline, including the India-based stockbroker Groww and the outage-tracking site Downdetector. After services were restored, Downdetector logged more than 4,500 Cloudflare-related reports.
With such a string of outages across key internet infrastructure providers, many businesses may start rethinking how reliant they’ve become on Cloudflare.
Steven Murdoch, a computer science professor at University College London, said the double outage raises important questions.
“After two outages in such a short time, people will start asking questions,” Murdoch said. “It’s too early to know if this is a systemic issue or just bad luck, but Cloudflare won’t be happy about it.”
Cloudflare, which markets itself on rock-solid reliability and high-performance cybersecurity, is used by companies to guard against attacks and boost site speed. However, recent failures — along with an Amazon Web Services outage in October affecting over 2,000 companies — have sparked a broader industry debate over the dangers of centralizing so much of the internet’s infrastructure.
“There’s a huge amount of centralisation,” Murdoch added. “Cloudflare has a great product, but that also creates vulnerabilities.”
Michał “rysiek” Woźniak, a DNS and infrastructure expert, said the latest incident highlights how fragile today’s hyper-centralized internet has become.
“This is the fourth major global outage since October 20,” he said.
Cloudflare says roughly 20% of the world’s websites use its services. It supports nearly 300,000 customers across 125 countries and blocks billions of cyberattacks daily. The company earns more than $500 million per quarter.
Woźniak says Cloudflare’s reliability-focused marketing is starting to look shaky.
“These companies have become too big to not fail,” he said. “And when they fail, the fallout is immediate and massive.”
Still, Murdoch noted there may be an unexpected silver lining for Cloudflare.
“When AWS went down, their share price actually rose because it showed how widely they’re used,” he said. “In some ways, outages like this highlight how essential Cloudflare really is.”
