Very Funny Tweet about Scaling
Yesterday I saw this very enjoyable tweet and it made me smile - quite a lot... because I do interviews often, and they are almost always about Architecture and Design, and scaling is one of the key use-cases to explore. It really helps see how the candidate approaches the problem:
Used to pay $5/mo on a small instance for my personal site. Then I discovered Kubernetes and realized my site didn't scale! No canary deployments! So I upgraded and pay $200/mo now. Took weeks to configure. Millions of people can now read my resume. Damn, it's never looked better
-- @malaroto
When faced with a scaling problem, AWS has been amazingly good - and amazingly bad for junior developers on the path of learning their craft. On the up-side, it's wonderfully full-featured, stable, available, on-demand in every way, and global. On the downside, it doesn't require any effort to use - so the first answer most jump to, is to build something that doesn't need to be built with unnecessary complexity that will work, but slows down the ability for someone else to understand the solution, and it's basically just wasting money.
"Not my dime" - is accurate, but not really the point. The point is to understand the problem, and then fix the problem, but if everything is able to be solved with millions of dollars of computing infrastructure, there seems to be no motivation to solve it with an hour of understanding and code refactoring.
So I giggle... they will have their day - a massive supercomputer in a grain of sand - ubiquitous and omnipresent... and then there will be no need to understand the why of issues... and that will be too bad. There will always be a need for craftsmanship.