In this tutorial series, we'll show you how to implement ArgoCD to manage and deploy applications to your Terraformed Kubernetes cluster running on the Linode Kubernetes Engine (LKE).

• Watch Terraforming Kubernetes on Linode or have some experience managing Kubernetes clusters
• Git installed
• Terraform installed
• Kubectl installed

git clone https://github.com/codingforentrepreneurs/terraforming-kubernetes-argocd
cd terraforming-kubernetes-argocd

Create an account on Linode and get an API Key in your linode account here.

Once you have a key, do the following:

echo "linode_api_token=\"YOUR_API_KEY\"" >> infra/terraform.tfvars

terraform -chdir=./infra init
terraform -chdir=./infra plan

If the plan looks good, run:

terraform -chdir=./infra apply

kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v1.7.0/deploy/static/provider/cloud/deploy.yaml

Or if you use helm:

helm upgrade --install ingress-nginx ingress-nginx \
  --repo https://kubernetes.github.io/ingress-nginx \
  --namespace ingress-nginx --create-namespace

Directly in the cert-manager docs:

kubectl apply -f https://github.com/cert-manager/cert-manager/releases/download/v1.11.0/cert-manager.yaml

apiVersion: cert-manager.io/v1
kind: ClusterIssuer
metadata:
  name: letsencrypt-prod
spec:
  acme:
    server: https://acme-v02.api.letsencrypt.org/directory
    email: [email protected]
    privateKeySecretRef:
      name: letsencrypt-prod-account-key
    solvers:
    - http01:
        ingress:
          class: nginx

Change [email protected] to your email. The http01 solver is great for non-wildcard domains. If you need a wildcard domain, you can consider using the DNS01 solver although that's a bit more complicated to setup.

ClusterIssuers do not care about namespaces as they are cluster-wide.

This manifest comes directly from the ArgoCD docs with one key change: the host. I used my domain name argocd.terraformingkubernetes.com but you can use any domain name you have control over.

My actual example is in config/ingress.yaml.

# https://argo-cd.readthedocs.io/en/stable/operator-manual/ingress/#kubernetesingress-nginx
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: argocd-server-ingress
  namespace: argocd
  annotations:
    kubernetes.io/ingress.class: nginx
    cert-manager.io/cluster-issuer: "letsencrypt-prod"
    nginx.ingress.kubernetes.io/force-ssl-redirect: "true"
    nginx.ingress.kubernetes.io/ssl-passthrough: "true"
spec:
  rules:
   - host: argocd.terraformingkubernetes.com
     http:
        paths:
        - path: /
          pathType: Prefix
          backend:
            service:
              name: argocd-server
              port:
                name: https
  tls:
    - hosts:
      - argocd.terraformingkubernetes.com
      secretName: argocd-secret # do not change, this is provided by Argo CD

Let's review the argocd-server manifest:

kubectl get deployment argocd-server -n argocd -o yaml

In here, we'll find a setting for ARGOCD_SERVER_INSECURE like so:

        env:
        - name: ARGOCD_SERVER_INSECURE
          valueFrom:
            configMapKeyRef:
              key: server.insecure
              name: argocd-cmd-params-cm
              optional: true

The value for this should be set to true so that ArgoCD is not redirecting to itself continously but instead letting the newly formed ingress work correctly.

kubectl patch configmap argocd-cmd-params-cm -n argocd -p '{"data":{"server.insecure":"true"}}'

Updating a configmap does not always trigger the deployment to restart. Let's do that now:

kubectl rollout -n argocd restart deployments/argocd-server

kubectl -n argocd get secret argocd-initial-admin-secret -o jsonpath="{.data.password}" | base64 -d

The argocd-repo-server, is responsible for polling git repos. We can change the timeout from the default of 3 minutes (3m) to 60 seconds (60s) or 10 days (10d). We get to pick how quickly this polling should happen.

kubectl get deployment argocd-repo-server -n argocd -o yaml

In the manifest, you'll see the following:

        env:
        - name: ARGOCD_RECONCILIATION_TIMEOUT
          valueFrom:
            configMapKeyRef:
              key: timeout.reconciliation
              name: argocd-cm
              optional: true

This tells us that the timeout.reconciliation key is declared in the ConfigMap (because of configMapKeyRef), named argocd-cm. With this in mind, let's update this default setting:

kubectl patch configmap argocd-cm -n argocd -p '{"data":{"timeout.reconciliation":"60s"}}'

As mentioned before, if we update a configmap it does not always trigger the deployment to updated as well. Let's do that now:

kubectl rollout -n argocd restart deployments/argocd-repo-server