You might have experienced the amount of time it takes and how error prone it is to bring up a few servers with operating systems and right software packages. Now, imagine thousands of locations , multiple servers in each location and bringing up all the software needed to create Kubernetes clusters at each location. Making a cluster out of servers requires multiple steps – configuring BIOS, installing operating systems, installing docker/K8s packages and installing basic network functions for connectivity. And the number of steps go up linearly with the number of locations. Hence, automation without physically going to each location is important. Then imagine Day2 operations – Knowing about health of clusters, @scale software upgrades – needed to deploy security fixes – and their complexity.