Several websites using Cloudflare were taken down yesterday due to bad software deployment as opposed to initial speculation that it was a cyber attack.
The outage started at 1:42 p.m. UTC and lasted for about 30 minutes. It affected visitors to websites like Shopify, Medium, Wattpad, and many others, who were shown 502 Bad Gateway errors.
The content delivery network service provider said the cause of the outage was the deployment of a single misconfigured rule within the Cloudflare Web Application Firewall (WAF) during a routine deployment of new Cloudflare WAF Managed rules.
The company explained that these new rules were deployed in order to improve its ability to block attacks that could bring a website down. Unfortunately, one of the rules contained a regular expression that led to a spike in CPU usage on Cloudflare’s system.
Websites experienced temporary paralysis
Cloudflare’s outage created a big hullabaloo as users of affected websites took to social media, particularly Twitter, to rant about the company’s miscalculation and create memes.
Aside from forcing websites to temporarily halt their operations, Cloudflare’s outage also
resulted in cryptocurrency and exchange websites displaying incorrect prices. Bitcoin news site, CoinDesk, was among those who tweeted about receiving wrong data from its providers due to the outage.
Cloudflare has acknowledged the damage it has caused its customers, and says it’s working on improving its processes.
“We recognize that an incident like this is very painful for our customers. Our testing processes were insufficient in this case and we are reviewing and making changes to our testing and deployment process to avoid incidents like this in the future.” – John Graham-Cumming, Chief Technology Officer, Cloudflare
Were you affected in any way by the internet outage that happened yesterday? Let us know in the comments below or over in our Facebook Group.