The cluster setup should work reliable now and servers will join after etcd is up. Furthermore, the client configuration will be added to the output directory now. Other files have seen improvements, too. But it is as it is, in heavy development :)
|6 months ago|
|output||6 months ago|
|templates||6 months ago|
|.gitignore||6 months ago|
|README.md||6 months ago|
|Vagrantfile||6 months ago|
|ansible.cfg||6 months ago|
|k3s_cluster.yml||6 months ago|
|k3s_container.yml||6 months ago|
|kubeconfig||6 months ago|
Ansible for k3s deployment
An Ansible repository to deploy a k3s kubernetes cluster.
Having different shapres of kubernetes on hand is somewhat mandatory for typical DevOps and development tasks.
This repository allows to build different types of k3s kubernetes setups locally and on specified machines.
This section describes the features of the repository.
You can set up the cluster, wherever you want. You just need to configure the inventory accordingly and you are good to go.
Your own machines
You can build different clusters by changing the amount of servers and agents. Some typical scenarios, you may be interested in, are described below.
A single kubernetes instance is useful for local development and testing. It is similar to microk8s or minikube setups.
You can setup a high available kubernetes cluster with 3 servers. This is useful to develop kubernetes services, ingress, HA testing and loadbalancer functionality.
Single server, multiple agents
If you want to test pod affinity, horizontal auto scaling or just want to use multiple agents, you can define one server and multiple agents.
HA Setup with multiple agents
Define at minimum 3 servers and 1 agent to get a high available setup. This is very close to a production cluster and can be useful to test production setups.
The repository will offer options to configure some additional scenarios.