do-bernetes
Deploy a Kubernetes cluster to Digital Ocean simply enough to learn about it in the process.
First of all, all attribution and inspiration for most of this project comes from kubernetes-digitalocean-terraform. This project is simplified for the sake of being an approachable way to experiment with Kubernetes.
Quick Start
If you are familiar with terraform and DigitalOcean, you can set the following environment variables, and run plan and apply to get started:
| Environment Variable | Description |
|---|---|
| TF_VAR_do_token | Your digital ocean API token |
| TF_VAR_ssh_fingerprint | $(ssh-keygen -E md5 -lf ~/.ssh/id_rsa.pub |
In this tutorial, we'll use DigitalOcean's API to help us manage droplets, disks, loadbalancers and more. You can get a DigitalOcean API token at: https://cloud.digitalocean.com/settings/tokens/new
Once you have your token setup, you can either paste it into the do_token field in deploy.tf, or you can export it to an environment variable with:
export TF_VAR_do_token=ENTER_YOUR_TOKEN_HERE
You will also need an SSH key uploaded to DigitalOcean. A guide for this can be found at How to Upload SSH Public Keys to a DigitalOcean Account
After you've done that, you'll need to find it's MD5 fingerprint, as DigitalOcean uses this to select which SSH key it uses to grant access to your droplets. To do so, you can run:
ssh-keygen -E md5 -lf ~/.ssh/id_rsa.pub | awk '{print $2}' | awk -F"MD5:" '{print $2}'
Once you have your SSH key fingerprint, you can export it to an environment variable by executing:
export TF_VAR_ssh_fingerprint=ENTER_YOUR_FINGERPRINT_HERE
Now, we're ready to run Terraform init, plan, and apply:
terraform init
terraform plan -out plan.out
terraform apply "plan.out"
This should bring up your Kubernetes cluster on DigitalOcean.
DigitalOcean controller manager and container storage interface
What follows is an explanation of how this project deploys the DigitalOcean cloud controller manager (CCM) and container storage interface (CSI) are deployed. Skip this if you just want to use Kubernetes or are disinterested in the details.
In order to give the CCM and CSI authorization to do this on your behalf, this project uses the TF_VAR_do_token variable you entered before.
It uses this variable to create a kubernetes secret. You can see the manifest in the secrets/ directory.
secret.yaml:
apiVersion: v1
kind: Secret
metadata:
name: digitalocean
namespace: kube-system
stringData:
access-token: "${TF_VAR_do_token}"
After this manifest is created, we deploy it to your kubernetes master by running:
kubectl create -f secret.yaml
With this access, CCM can provision loadbalancers, disks, and more through the DigitalOcean API.
Updating Versions
It is important to consult the cloud controller manager docs when updating to ensure that you use compatible versions of Kubernetes relative to the CCM. Compatibility here is no joke. See (https://github.com/digitalocean/digitalocean-cloud-controller-manager/blob/master/docs/getting-started.md#version) to make sure you use the right version.
kubectl apply -f https://raw.githubusercontent.com/digitalocean/digitalocean-cloud-controller-manager/master/releases/v0.1.5.yml
We also need to setup CSI that allows Kubernetes to provision disks through the API. This software also has tight coupling on version. To ensure for compatibility, see first: (https://github.com/digitalocean/csi-digitalocean#kubernetes-compatibility). DigitalOcean's CSI can deployed by running:
kubectl apply -f https://raw.githubusercontent.com/digitalocean/csi-digitalocean/master/deploy/kubernetes/releases/csi-digitalocean-v0.2.0.yaml
How to access your cluster
You can access your cluster by setting and exporting the KUBECONFIG variable. While in this project's top level directory, run:
export KUBECONFIG=$(pwd)/secrets/admin.conf
For confirmation, you should be able to now use the kubectl binary to see the status of your nodes:
kubectl get nodes
Example:
kubectl get nodes
NAME STATUS ROLES AGE VERSION
k8s-master Ready master 13m v1.13.2
k8s-worker-01 Ready <none> 10m v1.13.2