When we had to buy hardware to set up your tests, we were efficient as hell.
We would spend weeks, maybe even months, watching for deals and agonizing over cache, RAM, CPU, disk arrays, and storage size.
Someone still has to do that to make the implementation of clouds effective, but you don’t have to do it anymore. We barely have server rooms anymore.
Now you want to try out a virtual machine or build a resource group, you spin it up and go to town. For getting up and running quickly, this is a dream, no more weeks and months waiting for hardware to arrive.
It also means our implementations are less efficient, we don’t “throw” hardware at the problem anymore, we “throw” more “stuff” at it – more Virtual Machines, more public Ips, more networks.
And when things change, we don’t change them, we leave them the same.
This is the rise of the inefficient cloud, where tons of resources run on your environments that aren’t even touched or consumed, or squared to run correctly.
How do you fix this?
Give yourself some constraints, give yourself some containers, and hold yourself accountable for what you have done and what you are trying to change. Don’t let your code become inefficient because you haven’t set up a coordinated infrastructure.
Just because you can do it, doesn’t mean you should, make it better.
