Topline: Discord, Verizon, Shopify and more were down Tuesday morning, the second such large-scale outage in a week across the internet involving Cloudflare, a major Web services provider.
- Cofounder and CEO Matthew Prince tweeted Cloudflare was aware of the issue and said it was caused by a spike in CPU traffic processors.
- “The same thing that runs inside your phone or laptop is the same thing that runs our servers across 180 cities in the world. Our servers saw a massive spike in CPU. And just like your laptop can lock up when it sees a spike in [CPU processes], our primary, backup, and backup of backup systems had the CPU spike spill over to them, and it impacted most of our systems globally,” Prince said in an interview with Forbes.
- Prince made it clear the outage was not an attack by bad actors. “There’s no evidence this was attack-related,” he said.
- Services affected by the outage would have been unavailable 30 minutes or less, according to Prince. CloudFlare is working to determine the root cause of the outage.
- Down Detector, a website listing online services having issues in real time, was ironically brought down by the outage, so internet users were temporarily unable to report the services they couldn’t access.
Key background: Tuesday’s outage was different from last week’s, which was caused by a Verizon route leak. Altogether, three major outages have hit the internet since June.
Surprising fact: The outage caused CoinDesk to temporarily list incorrect cryptocurrency prices, including bitcoin for a mere $26. (Bitcoin is now correctly listed.)