Supergiant (SG) 2.x departs from the architecture of 1.x in drastic ways. Most simply speaking, not only is the code almost entirely rewritten, but each concern of SG is split out into its own "microservice." There is no longer "Supergiant." There are instead three distinct, independent applications.
For example, SG Capacity is no longer a function of SG (see the old documentation) that has to be turned on or off. It is a singular tool that can be run completely independent of SG.
If all a kube needs is a means of autoscaling its nodes, then SG Control (essentially the "Mothership") is not even needed. Users can install Capacity, use its API or UI, and forget about the rest.
The "SG K8s toolkit" is now, in reality, three separate applications. Even so, while these components can work separately, they can also work together in the same cluster, where each part contributes to the overall goal: making Kubernetes powerful and approachable, for anyone from developers to enterprises. This is what the architecture would look like when all three tools are running together:
A chart to show SG in its 4k, HDR, 121Hz, 9.3 surround sound, CRT, antibacterial glory!
Kubernetes, on its own, is an enterprise-grade tool. However, as users become more familiar with it, they find that there are still problems to solve. If these problems could be summed up with a list of questions, then each one is answered by an SG component:
- How do I make production-ready kubes more easily?
a. How do I do it on each cloud?
- How do I automate node scaling?
a. How do I do it intelligently?
- How do I best configure my apps and kube?
a. How can I automate this process?
- How do I coordinate all of these things at once?
"How do I make production-ready kubes more easily, and on each cloud I need to support?" Control empowers users with this ability. Learn more?
"How do I coordinate and manage all of these activities at once?" SG Control also helps with this. SG Control is like a "Mothership" or central hub through which SG Capacity and SG Analyze can be plugged in and controlled--of course, the tools still have their own, independent, UIs and APIs, if you choose to do things that way.
"How do I automate node-scaling in a way that I can reduce costs but without risking uptime?" SG Capacity can handle this complex process on its own. Learn more?
"How do I best configure my apps and kube without spending all my time on it?" SG Analyze helps teams run checks to keep an eye on their cluster, and provides recommendations for the best setup. Learn more?