Posted on July 3, 2019 at 12:33 PM
Cloudflare is one of the biggest global internet service providers. As such, any difficulty that the service might face tends to affect all of its users. It is not an exaggeration to say that Cloudflare’s issues tend to shake the entire internet, which is exactly what happened recently when the service saw technical issues for about an hour. At the same time, countless websites around the world went down, displaying an error message to everyone who tried to access them.
Cloudflare explained what happened
The issues did not last long, and the service resolved the issues within an hour, as mentioned. From that point on, everything continued working as usual. However, the entire incident is wrapped up in a sort of mystery, and many are wondering what had actually happened.
Of course, they did issue an official explanation, with its CTO, John Graham-Cumming, stating that the service’s visitors might have received 502 error messages for about 30 minutes. He stated that the issue came from a major growth of CPU utilization on their network, which was, in turn, caused by a ‘bad software deploy’ that was rolled back. After restoring things to what they were, Cloudflare’s issues were resolved.
Meanwhile, the company’s CEO, Matthew Prince, also explained that the bug caused Firewall process to start consuming CPU. At first, the service believed it was under attack, but shutting down the process restored the system to normal.
Not an attack?
It is clear that both officials are trying to point out that it was just an error, and that there was no attack that crippled their network. Prince also pointed out that this is not a Verizon Border Gateway Protocol network issue, which the service faced last week. Even so, many are still wondering how is it possible that such a small issue caused so many difficulties for the massive part of the internet?
The explanation is not that difficult to grasp. After all, Cloudflare operates a very popular CDN, which protects website owners from issues like comment spam attacks and peak loads, when working correctly. It is even capable of protecting them from DDoS attacks. But, when something is not working as it should, then issues like this one emerge.
The service’s CDN optimizes the delivery of website resources to websites’ visitors. In other words, it delivers visitors to the website’s static from its data centers, located all around the world. Meanwhile, the web servers used by websites only deliver dynamic content. In general, using Cloudflare’s network is a much faster way for visitors to reach sites, than for them to go to the site itself, directly.
As such, the CDN has grown to be rather popular, and it holds nearly 35% of the market, with Amazon’s CloudFront being the second most popular one, with 28.84% of the market. Cloudflare protects more than 16 million websites, including even some of the biggest and most popular ones, like Pinterest, BuzzFeed, Dropbox, and alike. That means that they will all experience issues if Cloudflare runs into a problem.
And this was quite a big one, with Prince even admitting that it was the biggest internal problem that the company has ever experienced. It did not only impact Cloudflare’s CDN, but also its DNS service.
But, the firm was rather transparent regarding the incident, and it did not shy away from taking the blame. Even so, the incident should be viewed as a lesson regarding how much the internet depends on a handful of major companies such as this one. Many see this as a big problem, but one which could be solved with P2P businesses and decentralization, which have grown quite popular in concept, especially in recent years.