This guide is for Coder v1.21.0 and later, which handle certificates differently from earlier versions of Coder. Ensure that you're reading the docs applicable to your Coder version.
This guide will show you how to install cert-manager v1.4.0 and set up your cluster to issue Let's Encrypt certificates for your Coder installation so that you can enable HTTPS on your Coder deployment. It will also show you how to configure your Coder hostname and dev URLs.
We recommend reviewing the official cert-manager documentation if you encounter any issues or if you want info on using a different certificate issuer.
You must have:
To add cert-manager to your cluster, run:
kubectl apply -f https://github.com/jetstack/cert-manager/releases/download/v1.4.0/cert-manager.yaml
More specifics can be found in the cert-manager install documentation.
Once you've started the installation process, verify that all the pods are running:
kubectl get pods -n cert-manager NAME READY STATUS RESTARTS AGE cert-manager-7cd5cdf774-vb2pr 1/1 Running 0 84s cert-manager-cainjector-6546bf7765-ssxhf 1/1 Running 0 84s cert-manager-webhook-7f68b65458-zvzn9 1/1 Running 0 84s
You can get the private key from the GCP Service Account using:
gcloud iam service-accounts keys create key.json \ --iam-account <service-account-name>@<project-name>.iam.gserviceaccount.com
The response should look similar to the following:
created key [44...3d] of type [json] as [key.json] for [<service-account-name>@<project-name>.iam.gserviceaccount.com]
Next, configure the cluster issuer secret, and add it to cert-manager's namespace:
kubectl -n cert-manager create secret generic \ clouddns-dns01-solver-svc-acct --from-file=./key.json
If successful, you'll see a response similar to:
Using the text editor of your choice, create a new
letsencrypt.yaml (you can name it whatever you'd like) that includes
your newly created private key:
apiVersion: cert-manager.io/v1 kind: ClusterIssuer metadata: name: letsencrypt spec: acme: privateKeySecretRef: name: gclouddnsissuersecret server: https://acme-v02.api.letsencrypt.org/directory solvers: - dns01: cloudDNS: # The ID of the GCP project project: <project-id> # This is the secret used to access the service account serviceAccountSecretRef: name: clouddns-dns01-solver-svc-acct key: key.json
More information on the values in the YAML file above can be found in the dns01 solver configuration documentation.
Apply your configuration changes:
kubectl apply -f letsencrypt.yaml
If successful, you'll see a response similar to:
Note: If you are providing an ingress, certificates can be automatically created with an ingress annotation. See the cert-manager docs for details. If you are unsure whether you are using an ingress or not, continue with this step.
In a text editor, create a new file called certificate.yaml and paste the following:
apiVersion: cert-manager.io/v1 kind: Certificate metadata: name: coder-certs namespace: coder # Your Coder deployment namespace spec: commonName: "*.coder.example.com" dnsNames: - "coder.example.com" - "*.coder.example.com" issuerRef: kind: ClusterIssuer name: letsencrypt secretName: coder-certs
Be sure to change
coder.example.com to the domain for your Coder deployment.
While this example uses a single domain, a separate domain can be created for
dev URLs or even omitted if you do not have
dev URLs enabled.
Once you're done, deploy the certificates.
kubectl apply -f certificate.yaml
At this point, you're ready to install Coder. However, to use all of the functionality you set up in this tutorial, use the following command instead:
helm upgrade --install coder coder/coder --namespace coder \ --version=<CODER_VERSION> \ --set coderd.devurlsHost="*.coder.example.com" \ --set coderd.tls.devurlsHostSecretName="coder-certs" \ --set coderd.tls.hostSecretName="coder-certs" \ --wait
The cluster-issuer will create the certificates you need, using the values
provided in the
helm install command for the dev URL and host secret.
There are additional steps to make sure that your hostname and Dev URLs work.
Check the contents of your namespace
kubectl get svc -n <your_namespace> -o wide
Find the service/coderd line, and copy the external IP value shown.
Return to Google Cloud Platform, navigate to the Cloud DNS Console, and select the Zone that your cluster is in.
Note: You will need to create two A records, one for both the hostname and Dev URLs
Click Add Record Set
Provide your DNS Name
a. If you're configuring the hostname, this value will be a standard domain
b. If you're configuring your dev URLs, this will be a wildcard domain (e.g.,
Set the Resource Record Type to A
Copy and paste the external IP address associated with the service/coderd
line from your terminal to the
IPv4 Address field.
At this point, you can return to step 6 of the installation guide to obtain the admin credentials you need to log in.
If you are not getting a valid certificate after redeploying, see cert-manager's troubleshooting guide for additional assistance.