Most people don’t care about content delivery networks or know anything about them. They would merrily go about their day visiting websites not caring about how they ended up on their device. On Tuesday this week the world suddenly had a hard and fast wake up call on how they worked when half the internet appeared to be taken out for 40mins and nobody could get work done. I’m guessing that if you’re a user of github.com, twitch or gov.uk you probably weren’t aware they all depended on Fastly. This
bbc article covers things in more detail, but ultimately lays the blame on a single customer changing some settings.
One customer, one company…half the internet. That’s a lot of productive man hours down the drain due to a huge single point of failure.
Of course as developers we all follow the old adage of “minimising our dependencies” and would never make use of such services… Ahem. The reality is that rolling our own CDN’s would likely result in more downtime and less productivity than 40 mins every once in a blue moon.
In another article this issue, github highlights the open source libraries used by top tech companies that like Fastly, are invisible to most people.
The stark reality is without services and tools like this many web applications we depend upon simply wouldn’t exist.
In other news, I’ve enjoyed a week away from everything over half term and relaxed by the beach. We seem to be in pretty much the same position lockdown wise as last summer, except this time we’re armed with a vaccine.
If you have any comments on this issue or content suggestions for the future, please feel free to reply to this email and let me know.
Until next time, keep on shipping!
Ian
