This week’s post was actually going to be about something different, but as you know, I like to respond to certain trending tech news so here we are.
You may have seen in the news that on Tuesday 8th June, a number of big corporate websites briefly went offline or suffered downtime.
Among these websites were Amazon, PayPal, GitHub and the UK Government (gov.uk).
So what went wrong and why did it affect so many different websites?
Well, it turns out that a popular CDN provider called Fastly, suffered an outage, which was being used as part of all of these websites’ infrastructure.
But what is a CDN?
A CDN stands for Content Delivery Network and is responsible for serving you these websites at the fastest pace possible (among other things).
Websites not using a CDN are usually hosted on 1 server in 1 location e.g. a datacenter in the USA.
Websites using a CDN are hosted on multiple “edge” servers, usually around the world e.g. UK, USA, France, Germany etc…, so that when you visit such website, you will be served the copy of the website stored on the server closest to your location e.g. the UK one if you live in the UK, resulting in a faster experience.
All of the “edge” servers are interconnected by a high-speed backbone network and supported by machines called “load balancers”, which decide what server will serve you the website and prevent the overloading of 1 particular server.
CDNs are often used by larger companies, since they provide redundancy i.e. if 1 server goes down or is too full, the “load balancer” can just choose the next closest server, but also because they prevent DDoS attacks and too much traffic slowing the website down.
Fastly CDN’s blunder on Tuesday was actually a rare instance of an entire CDN going down, which happened to be fixed by Fastly within 45 minutes and I’m sure they have learned from the error.
So there you go, I hope you enjoyed this post as much as I did writing it.
Thanks for reading!