*The above claim depends on your internet and processing power. ^-^
This repository exists only to get rid of the phrase, "But it works in my pc!" The cluster bootstrap mechanism currently supports debian based releases preferably Ubuntu. Windows and Mac docker desktop based installation is under consideration.
Tool | Version |
---|---|
Docker | 25.0.3 |
Kubectl | 1.29.2 |
KinD | 0.21.0 |
Helm | 3.14.2 |
Helmfile | 0.162.0 |
Calico | 3.25.0 |
Nfs-subdir-external-provisioner | 4.0.18 |
Elasticsearch | 8.12.2 |
Kibana | 8.12.1 |
Fluent Bit | 2.2.2 |
Prometheus | 2.49.1 |
Grafana | 10.3.3 |
Metric Server | 0.7.0 |
Make needs to be installed
sudo apt update
sudo apt install make
Running this for the first time and don't have docker, kubectl, kind, helm, helmfile installed?
make first
Install the cluster with all components if you have docker, kubectl, kind, helm, helmfile installed already.
make up
Destroy the cluster and nfs configuration. This will not delete nfs shared directory contents.
make down
Important: Make sure to update
server
innfs.yaml
underdev/values
directory. If you are not sure, reinstall the cluster withmake down up
Optional: Point to this cluster's kubeconfig
export KUBECONFIG=~/.kube/config
If the applications are already available as a helm chart, add them in helmfile.yaml
and run,
make apply
Alternatively apply application manifests with kubectl,
kubectl apply -f <your_yaml_manifest_file_or_directory>
Info: Port 30000-30020 and 32767 are already open. You can open more in
cluster/kind-config.yaml
as well.
Get the elastic
user password to access Kibana http://localhost:30000/
kubectl get secrets/elasticsearch-kibana -n platform --template='{{ index .data "kibana-password" | base64decode}}'
Get the admin
user password to access Grafana http://localhost:30001/
kubectl get secrets/prometheus-grafana -n platform --template='{{index .data "admin-password" | base64decode}}'
Prometheus dashboard at http://localhost:30002/
Alertmanager dashboard at http://localhost:30003/