Skip to main content

Run Dagger on Kubernetes

Introduction​

This guide outlines how to run, and connect to, the Dagger Engine on Kubernetes.

Assumptions​

This guide assumes that you have:

  • The Dagger CLI installed locally.
  • Helm v3.x available locally.
  • A running Kubernetes cluster (tested with Kubernetes v1.28).

Step 1: Deploy Dagger with Helm​

Deploy Dagger on your Kubernetes cluster with Helm:

helm upgrade --install --namespace=dagger --create-namespace \
    dagger oci://registry.dagger.io/dagger-helm

Wait for the Dagger Engine to become ready:

kubectl wait --for condition=Ready --timeout=60s pod \
    --selector=name=dagger-dagger-helm-engine --namespace=dagger

You can find more information on what was deployed using the following command:

kubectl describe daemonset/dagger-dagger-helm-engine --namespace=dagger

Step 2: Connect Dagger CLI to Dagger Engine pod​

Get a Dagger Engine pod name:

DAGGER_ENGINE_POD_NAME="$(kubectl get pod \
    --selector=name=dagger-dagger-helm-engine --namespace=dagger \
    --output=jsonpath='{.items[0].metadata.name}')"
export DAGGER_ENGINE_POD_NAME

Next, set the _EXPERIMENTAL_DAGGER_RUNNER_HOST variable so that the Dagger CLI knows to connect to the Dagger Engine that you deployed as a Kubernetes pod:

_EXPERIMENTAL_DAGGER_RUNNER_HOST="kube-pod://$DAGGER_ENGINE_POD_NAME?namespace=dagger"
export _EXPERIMENTAL_DAGGER_RUNNER_HOST

Finally, run an operation that shows the kernel info of the Kubernetes node where this Dagger Engine runs:

dagger query <<EOF
{
    container {
        from(address:"alpine") {
            withExec(args: ["uname", "-a"]) { stdout }
        }
    }
}
EOF

This is what a successful response should look like:

┣─╮
│ ▽ init
│ █ [0.64s] connect
│ ┣ [0.52s] starting engine
│ ┣ [0.12s] starting session
│ ┃ OK!
│ ┻
█ [2.44s] dagger query
┣ [0.00s] loading module
█ [2.44s] query
┃ {
┃     "container": {
┃         "from": {
┃             "withExec": {
┃                 "stdout": "Linux buildkitsandbox 6.1.0-12-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.52-1 (2023-09-07) x86_64 Linux\n"
┃             }
┃         }
┃     }
┃ }
┣─╮
│ ▽ from alpine
│ █ [1.26s] resolve image config for docker.io/library/alpine:latest
│ █ [0.18s] pull docker.io/library/alpine:latest
│ ┣ [0.03s] resolve docker.io/library/alpine@sha256:34871e7290500828b39e22294660bee86d966bc0017544e848dd9a255cdf59e0
│ ┣ [0.61s] ███████████████████ sha256:c926b61bad3b94ae7351bafd0c184c159ebf0643b085f7ef1d47ecdc7316833c
│ ┣ [0.18s] extracting sha256:c926b61bad3b94ae7351bafd0c184c159ebf0643b085f7ef1d47ecdc7316833c
│ ┣─╮ pull docker.io/library/alpine:latest
│ ┻ │
█◀──╯ [0.24s] exec uname -a
┃     Linux buildkitsandbox 6.1.0-12-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.52-1 (2023-09-07) x86_64 Linux
┻
• Engine: dagger-dagger-helm-engine-bvbtk (version v0.9.3)
⧗ 3.10s ✔ 12

The line above starting with Engine: confirms the Dagger Engine that the CLI connected to.

To double-check that the operations are running on the Kubernetes cluster, follow the pod logs:

kubectl logs pod/$DAGGER_ENGINE_POD_NAME --namespace=dagger --follow

Conclusion​

This guide demonstrated the simplest approach to using Dagger on Kubernetes. For more complex scenarios, such as setting up a Continuous Integration (CI) environment with Dagger on Kubernetes, use the following resources:

info

If you need help troubleshooting your Dagger deployment on Kubernetes, let us know in Discord or create a GitHub issue.