Tear-down Live Hosts
We now come to the first applicable teardown script. This script undoes all the
work that the
standup-live script does and returns hosts to a vanilla state.
This is important to have if we plan on changing the host plan.
And it has the option for tearing down individual live hosts, without tearing them all down:
apb Hosts/teardown-live.yml -e "killhosts=pi4,pi5"
You probably only need to do this if you need low-security host access restored, and don't want to re-image, the host is in need of deep maintenance, or if you want to re-organize your cluster roles. Be it unlikely, it is also essential for a clean host exit in the case of exercising a Hydra Option.
Further, if you are running this command to permanently remove a host from the
cluster, you will need to comment out that hosts entry on your host plan by
# in front, before running higher commands that rely on all hosts.
Hosts Are Finished
Taking a step back, we now hopefully have a multi-architecture, multi-manufacturer, multi-supply-chain, multi-OS host array. It's not wholly unlikely that one component is compromised or on the verge of breaking, it's very unlikely that they all are.
You made it. That's it for hosts, from here on out, it's all higher level software, staring with the big man on campus, Kubernetes. Get excited!