Database "scaleoutability" has historically been a very complex challenge to tackle. And usually it requires some form of sharding that while it solves some problem it add an entire set of new challenges. IF you have struggled deciding the correct sharding key and strategy you know what I mean. Well, no more! With Named Replicas creating applications that can elastically scale immensely is quite easy. All you need to do is challenge yourself and will to make a very small architectural change that will bring you a huge advantage in terms of flexibility and performance. Let's take a look together at Named Replicas and Tag-Based query routing and how you can fit them in your solution.