Honestly, in terms of pure ROI and operational load, it doesn't look like the battery life in your current data center is the right solution for this problem.
From the operator’s point of view, a container system makes sense when there are rental and heterogeneity problems that are difficult to solve with VM / VPSes, and a plant has at least dozens of nodes in size. Launching any container infrastructure in HA mode is a big job, and there are many corner cases that require special attention of specialists. The need must be large enough to make sense to make these investments.
The plant, as described, with redundancy, can run on dozens of virtual machines / VPS. Achieving the desired accessibility levels requires a thorough architecture, but stateless database and application management templates on HA virtual machines with, say, 3x scalability, are pretty well established.
There are still many discoveries in the container world. Especially with Kubernetes, every quarter there is a completely new release with new open corners.
Of course, it is really interesting to know about it, but he is still in a state where it is wonderful to see that he is working, and not bored.
source share