Apparently not wanting GoDaddy and GitHub to have all the outage fun this week, Cloudflare confirmed on Twitter that it had “issues” this morning. Many users found their sites unavailable, and some sites may still be experiencing trouble. However, I’m hearing that not all users of the service were affected.
An upstream provider issue caused issues for sites across our network this morning. Should be fixed now and clearing.Our sincere apologies
— CloudFlare (@CloudFlare) September 15, 2012
Cloudflare provides security for websites by acting as a sort of firewall between users and the servers where sites are actually hosted. This allows it to filter malicious traffic before it ever hits its target. It also means that if the Cloudflare servers go down, so do the sites they protect.
The company has experienced explosive growth. “Collectively, the startup now sees 70 billion page views per month and 600 million uniques — more traffic per month than Instagram, eBay, Amazon, Aol, Apple and many more, combined,” Rip wrote earlier this week.
It’s done an amazing job scaling quickly, and the outage was relatively brief. But what we’ve been seeing with the outages this week, and what we see whenever other high-profile services like Amazon Web Services go down, is just how many other sites and services can be affected by an outage at a single company.
Update: See comments for an explanation from Cloudflare.