FP ExplainersJun 22, 2022 11:21:56 IST
Last night, nearly half of all websites on the internet were inaccessible and were showing error 500. Error 500 mainly occurs when there is an issue with the CDN or content distribution networks.
As a result, some popular services and websites like Discord, League of Legends, a number of popular news websites and some of Google’s services were interrupted.
Large chunks of the internet are returning an nginx 500 Internal Server Error right now (I’m assuming something’s messed up at Cloudflare or similar) but this is definitely my favourite pic.twitter.com/cK5x49ZLiR
— Neil Studd (@neilstudd) June 21, 2022
Yesterday’s internet blackout was mainly caused due to an outage of Cloudfare’s servers. The online security company was quick to identify the problem and rectify the issue, getting these services back up and running, but it was a chaotic few minutes there.
More Sites Reporting
❌ Google Services
❌ ARK Survival Evolved
— GhostfromTexas (@GhostfromTexas) June 21, 2022
According to Cloudflare’s incident report, the issue was reported at approximately 6:34 a.m. UTC, the problem was identified just over 20 minutes later around 6:57 a.m., and the fix was implemented at around 7:20 a.m.
Some major services that were interrupted included Amazon Web Services and Google Services. Anyone familiar with the inner workings of today’s internet will realise that nearly 60 to 70 per cent of the entire internet is hosted on Amazon’s servers and that AWS handles over 75 per cent of the Internet’s existing traffic on any single day.
Any disruptions that effects AWS and its subsidiaries will disrupt a major chunk of the internet.
The crypto ecosystem in particular was hit badly. Nearly all of the major crypto wallets and exchanges use services from Cloudflare.
Gotta love seeing 95% of the crypto ecosystem down because of a Cloudflare outage
Very decentralized! Great work guys!
— Altcoin Psycho (@AltcoinPsycho) June 21, 2022
Platforms like Coinbase uses the network, proxy & security provided by Cloudflare. As a result, the crypto market crashed even further, with several users unable to carry out transactions.
Anonymous sources inside Cloudflare believe that the reason why their servers went down was that they were trying to install an update patch to their server management systems, which did not go according to plan and caused an error.
Although Cloudflare was quick to realise this and promptly took action to rectify the problem, it just goes on to show, how badly the modern internet infrastructure has been set up that one company’s server going down somehow results in basically 1/3 of the internet going down.