Before proceeding, please ensure that you have a Kubernetes cluster running K8s 1.19+ and have Helm 3.5+ installed.

You'll also want to install the latest version of Coder locally in order to log in and manage templates.

Coder supports two release channels: mainline for the true latest version of Coder, and stable for large enterprise deployments. Before installing your control plane via Helm, please read the Releases document to identify the best-suited release for your team, then specify the version using Helm's --version flag.

The version flags for both stable and mainline are automatically filled in this page.

If you need help setting up k8s, we have a repo with Terraform configuration to provision Coder on Google GKE, Azure AKS, AWS EKS, DigitalOcean DOKS, IBMCloud K8s, OVHCloud K8s, and Scaleway K8s Kapsule.

Install Coder with Helm

  1. Create a namespace for Coder, such as coder:

    kubectl create namespace coder
  2. Create a PostgreSQL deployment. Coder does not manage a database server for you.

    If you're in a public cloud such as Google Cloud, AWS, Azure, or DigitalOcean, you can use the managed PostgreSQL offerings they provide. Make sure that the PostgreSQL service is running and accessible from your cluster. It should be in the same network, same project, etc.

    You can install Postgres manually on your cluster using the Bitnami PostgreSQL Helm chart. There are some helpful guides on the internet that explain sensible configurations for this chart. Example:

    # Install PostgreSQL
    helm repo add bitnami https://charts.bitnami.com/bitnami
    helm install coder-db bitnami/postgresql \
        --namespace coder \
        --set auth.username=coder \
        --set auth.password=coder \
        --set auth.database=coder \
        --set persistence.size=10Gi

    The cluster-internal DB URL for the above database is:

    postgres://coder:[email protected]:5432/coder?sslmode=disable

    Ensure you set up periodic backups so you don't lose data.

    You can use Postgres operator to manage PostgreSQL deployments on your Kubernetes cluster.

  3. Create a secret with the database URL:

    # Uses Bitnami PostgreSQL example. If you have another database,
    # change to the proper URL.
    kubectl create secret generic coder-db-url -n coder \
       --from-literal=url="postgres://coder:[email protected]:5432/coder?sslmode=disable"
  4. Add the Coder Helm repo:

    helm repo add coder-v2 https://helm.coder.com/v2
  5. Create a values.yaml with the configuration settings you'd like for your deployment. For example:

      # You can specify any environment variables you'd like to pass to Coder
      # here. Coder consumes environment variables listed in
      # `coder server --help`, and these environment variables are also passed
      # to the workspace provisioner (so you can consume them in your Terraform
      # templates for auth keys etc.).
      # Please keep in mind that you should not set `CODER_HTTP_ADDRESS`,
      # they are already set by the Helm chart and will cause conflicts.
              # You'll need to create a secret called coder-db-url with your
              # Postgres connection URL like:
              # postgres://coder:password@postgres:5432/coder?sslmode=disable
              name: coder-db-url
              key: url
        # (Optional) For production deployments the access URL should be set.
        # If you're just trying Coder, access the dashboard via the service IP.
        - name: CODER_ACCESS_URL
          value: "https://coder.example.com"
      #  secretNames:
      #    - my-tls-secret-name

    You can view our Helm README for details on the values that are available, or you can view the values.yaml file directly.

  6. Run the following command to install the chart in your cluster.

    For the mainline Coder release:

    helm install coder coder-v2/coder \
        --namespace coder \
        --values values.yaml \
        --version 2.12.1

    For the stable Coder release:

    helm install coder coder-v2/coder \
        --namespace coder \
        --values values.yaml \
        --version 2.11.2

    You can watch Coder start up by running kubectl get pods -n coder. Once Coder has started, the coder-* pods should enter the Running state.

  7. Log in to Coder

    Use kubectl get svc -n coder to get the IP address of the LoadBalancer. Visit this in the browser to set up your first account.

    If you do not have a domain, you should set CODER_ACCESS_URL to this URL in the Helm chart and upgrade Coder (see below). This allows workspaces to connect to the proper Coder URL.

Upgrading Coder via Helm

To upgrade Coder in the future or change values, you can run the following command:

helm repo update
helm upgrade coder coder-v2/coder \
  --namespace coder \
  -f values.yaml

Kubernetes Security Reference

Below are common requirements we see from our enterprise customers when deploying an application in Kubernetes. This is intended to serve as a reference, and not all security requirements may apply to your business.

  1. All container images must be sourced from an internal container registry.

  2. All containers must run as non-root user

  3. Containers cannot run privileged

    • Coder's control plane does not run as privileged. We disable allowPrivilegeEscalation by default.
    • Workspace pods do not require any elevated privileges, with the exception of our envbox workspace template (used for docker-in-docker workspaces, not required).
  4. Containers cannot mount host filesystems

    • Both the control plane and workspace containers do not require any host filesystem mounts.
  5. Containers cannot attach to host network

    • Both the control plane and workspaces use the Kubernetes networking layer by default, and do not require host network access.
  6. All Kubernetes objects must define resource requests/limits

    • Both the control plane and workspaces set resource request/limits by default.
  7. All Kubernetes objects must define liveness and readiness probes

    • Control plane - The control plane Deployment has liveness and readiness probes configured by default here.
    • Workspaces - the Kubernetes Deployment template does not configure liveness/readiness probes for the workspace, but this can be added to the Terraform template, and is supported.

Load balancing considerations


If you are deploying Coder on AWS EKS and service is set to LoadBalancer, AWS will default to the Classic load balancer. The load balancer external IP will be stuck in a pending status unless sessionAffinity is set to None.

    type: LoadBalancer
    sessionAffinity: None

AWS recommends a Network load balancer in lieu of the Classic load balancer. Use the following values.yaml settings to request a Network load balancer:

    externalTrafficPolicy: Local
    sessionAffinity: None
    annotations: { service.beta.kubernetes.io/aws-load-balancer-type: "nlb" }

By default, Coder will set the externalTrafficPolicy to Cluster which will mask client IP addresses in the Audit log. To preserve the source IP, you can either set this value to Local, or pass through the client IP via the X-Forwarded-For header. To configure the latter, set the following environment variables:

      value: X-Forwarded-For
      value: # this will be the CIDR range of your Load Balancer IP address


In certain enterprise environments, the Azure Application Gateway was needed. The Application Gateway supports:

  • Websocket traffic (required for workspace connections)
  • TLS termination


You can view Coder's logs by getting the pod name from kubectl get pods and then running kubectl logs <pod name>. You can also view these logs in your Cloud's log management system if you are using managed Kubernetes.

Kubernetes-based workspace is stuck in "Connecting..."

Ensure you have an externally-reachable CODER_ACCESS_URL set in your helm chart. If you do not have a domain set up, this should be the IP address of Coder's LoadBalancer (kubectl get svc -n coder).

See troubleshooting templates for more steps.

Next steps

See an opportunity to improve our docs? Make an edit.