This cluster is designed to allow quick prototyping of k8s.
It is deployed and configured using Ansible to allow a number of base level configurations.
- CNI Networking
- metallb loadbalancers
- Grafana and Prometheus Monitoring
- Traefik V2, socat proxy forwarders for metallb
- Nginx, socat proxy forwarders for metallb
To ensure python dependencies.
pip3 install -r requirements.txt --useransible-galaxy collection install -r requirements.yml| Variables | Default | Description |
|---|---|---|
| cluster_reset | False | This will delete and re-install the kind cluster |
| docker_config | True | Inject docker config from user into cluster |
| kind_workers | 3 | Number of Kind Worker Nodes |
| external_api | False | Allow access to the k8s api via 0.0.0.0 Security Goose says this is dangerous use at your own peril. |
| kubernetes_version | "" | Use Kind image to define k8s image. |
| cni_network | calico | Set to either "" or calico for installing network |
| metallb_enabled | True | Metallb Load Balancer |
| ingress_enabled | True | Configure a Ingress |
| ingress_type | traefik | Nginx/Traefik |
| monitoring_enabled | False | Prometheus installation for testing monitoring |
ansible-playbook start-local-cluster.yml
ansible-playbook start-local-cluster.yml -t clean- Add additional CNI providers.
- Allow configuring of versions for various services, eg. ingress, monitoring.
- Inject values into config for ingress, monitoring helm charts.
- Make configuration more plugable.
Outside of creating the above basic configurations no other features and or functionality is expected to be added. This project is configured to allow a flexible development environment for developing in kubernetes and allow for the environment to be destroyed and recreated quickly and easily.
It is not meant for any production workloads.