'Major' Cloudflare Outage Briefly Knocks Much Of Internet Offline

‘Major outage impacted all Cloudflare services globally,’ the company wrote on its website. ‘We saw a massive spike in CPU that caused primary and secondary systems to fall over.’

Cloudflare told customers Tuesday that it experienced "major" networkwide issues that briefly took websites around the world offline.

"Aware of major @Cloudflare issues impacting us network wide," Cloudflare co-founder and CEO Matthew Prince wrote on Twitter at 10:08 a.m. ET Tuesday. "Team is working on getting to the bottom of what’s going on. Will continue to update."

Prince said in a tweet 14 minutes later that Cloudflare appears to have mitigated the issue causing the outage and has restored traffic. The San Francisco-based website performance company is currently working to restore all services globally, Prince wrote on Twitter, and will share more details once the company has them.
[Related: New Cloudflare Channel Chief Unveils ‘Partner Program 2.0’]

Cloudflare said in a 9:52 a.m. ET message on its website that it was observing network performance issues, causing customers to experience 502 Bad Gateway errors while accessing sites on Cloudflare. Cloudflare serves as the backbone for much of the internet and keeps services like Discord, Marketo and DigitalOcean up and running.

"Major outage impacted all Cloudflare services globally," the company wrote on its System Status website. "We saw a massive spike in CPU that caused primary and secondary systems to fall over."

Cloudflare said it shut down the process that was causing the CPU spike and was able to restore service to normal within 30 minutes. At 10:57 a.m. ET, Cloudflare said it had resolved the issue and services have resumed normal operation. The company said it is investigating the root cause of what happened. Cloudflare didn't immediately respond to a request for additional comment.

As of 10:50 a.m. ET, Cloudflare said its edge network and dashboard/API management service is being re-routed in 10 locations globally: Houston; Montgomery, Ala.; Rio de Janeiro, Brazil; Belgrade, Serbia; Baghdad, Iraq; Riyadh, Saudi Arabia; Karachi, Pakistan; Chennai, India; Mumbai, India; and Perth, Australia.

Customers have been reporting issues with Cloudflare since 9:56 a.m. ET today, according to downdetector.com, with the number of outage reports peaking at 3,227 right around 10 a.m. ET. An outage map on downdetector.com indicates significant problems with Cloudflare in New York, Los Angeles, Detroit, Minneapolis, Denver, Dallas, Houston and Miami.

Cloudflare users took to Twitter to complain about the outage under the hashtag #CloudflareDown.