This is so easy and simple it makes me nuts:
At this point you have a 3 node deployment of rancherOS with rancher server running on rancher-01 and rancher agent(s) running on nodes rancher-02 and rancher-03.
If you’re an enterprise class user:
- then you might need to increase the number of nodes in the Vagrantfile, redeploy, relink all of the agents and so on
- you’ll want to deploy your own registry server
- and you might want to disable or intercept the public repos via some dns trickery
The rancher tooling does not include any OS or rancher tools for monitoring, health, alerting, logging and so on… so you really have to start bolting IoT (internet of things) onto your environment. It also depends on whether you want to self host these operations or add SaaS type operations.
One thing to note about any of these orchestration applications. (a) there is a certain amount of magnification (or Mandelbrot) that takes place between bare metal and containers; so keep an eye on the APIs because they may be the one constant. (b) you will still need bare metal servers to provide services that you do not want to be inside the VM or container. HA/load balancers, NTP, DNS, PXE server and storage servers come to mind.
PS: if you are going to run rancherVMs inside your containers then you gotta have huge bare metal hosts and you need a way to orchestrate the work as rancher leaves that to the user.