Recently, I’ve noticed a troubling anti-pattern when building software.
That anti-pattern is looking to solve the scaling problems, before you’ve even managed to build anything.
While it’s important to note the limitations you may face, it’s even more important to deliver working software that achieves your goals as quickly and as well as possible.
Let your scaling problems be scaling problems, don’t spend countless hours, days, weeks and months debating whether you’ll be able to support 1000 micro services on your platform before you even have the first one out in front of customers.
Getting that first thing out there allows you to then learn from, make adjustments to and nail your approach for the next one. Keep that loop going and the scaling problems will show themselves when they actually start to become problems. You may even find that you change direction completely, nullifying the problem you were debating in the first place.