gke-deploy

packaged by Chainguard

Chainguard Container for gke-deploy

gke-deploy prepares and applies Kubernetes manifests to Google Kubernetes Engine clusters, expanding them with recommended labels before applying and waiting for readiness. Bundled with the gcloud SDK, kubectl and the GKE auth plugin it shells out to, plus kustomize for its documented pipe.

Chainguard Containers are regularly-updated, secure-by-default container images.

Download this Container Image

For those with access, this container image is available on cgr.dev:

docker pull cgr.dev/ORGANIZATION/gke-deploy:latest

Be sure to replace the ORGANIZATION placeholder with the name used for your organization's private repository within the Chainguard Registry.

Compatibility Notes

This image replaces the upstream gcr.io/cloud-builders/gke-deploy build step for running gke-deploy itself, with the differences below. The entrypoint is /bin/gke-deploy, so arguments passed to docker run reach the CLI unchanged, and the same subcommands are available.

The image runs as nonroot, so $HOME is /home/nonroot, which that user owns and can write. The /root/.config volume is still declared, inherited from upstream, but nothing under UID 65532 can write to it because /root belongs to root; treat it as vestigial rather than a place to mount SDK state. The working directory is /, which UID 65532 cannot write either, so give prepare and run an explicit --output rather than relying on its ./output default. Make any mounted output directory writable by UID 65532, and if you override the user to a UID that is not in /etc/passwd, set HOME and CLOUDSDK_CONFIG yourself, because the runtime then falls back to /.

Alongside gke-deploy the image carries the tools it shells out to, gcloud, gsutil and kubectl, along with the GKE auth plugin that kubectl loads, plus kustomize for its documented pipe. yq, envsubst, nomos and local-extract, which upstream also installs, are not included.

Getting Started

Print the available subcommands:

docker run --rm cgr.dev/ORGANIZATION/gke-deploy:latest --help

Expand manifests without contacting a cluster. prepare adds the recommended Kubernetes labels and writes both a suggested and an expanded copy:

docker run --rm -u "$(id -u):$(id -g)" -v "$PWD:/work" cgr.dev/ORGANIZATION/gke-deploy:latest \
  prepare -f /work/configs -a hello -o /work/out

Finished preparing deployment.

The expanded manifests land in out/expanded/aggregated-resources.yaml, carrying the app.kubernetes.io/managed-by: gcp-cloud-build-deploy label that marks them as deployed by this tool.

Apply manifests to a cluster, using a kubeconfig that carries its own credentials:

docker run --rm -u "$(id -u):$(id -g)" -v "$PWD:/work" \
  -e KUBECONFIG=/work/kubeconfig -e HOME=/tmp \
  cgr.dev/ORGANIZATION/gke-deploy:latest \
  apply -f /work/configs -n default --timeout 300s

apply waits up to --timeout for the workload to become ready, so a non-zero exit means either the apply itself failed or the rollout did not converge in time.

A kubeconfig generated by gcloud container clusters get-credentials does not carry credentials. It delegates to gke-gcloud-auth-plugin, which reads the gcloud configuration, so for a GKE cluster also mount an authenticated config and point CLOUDSDK_CONFIG at it:

docker run --rm -u "$(id -u):$(id -g)" -v "$PWD:/work" -v "$HOME/.config/gcloud:/gcloud:ro" \
  -e KUBECONFIG=/work/kubeconfig -e CLOUDSDK_CONFIG=/gcloud -e HOME=/tmp \
  cgr.dev/ORGANIZATION/gke-deploy:latest \
  apply -f /work/configs -n default --timeout 300s

In Google Cloud Build, use the image in place of the upstream build step:

steps:
  - name: cgr.dev/ORGANIZATION/gke-deploy:latest
    env:
      - HOME=/tmp
      - CLOUDSDK_CONFIG=/tmp/gcloud
    args:
      - run
      - --filename=configs/
      - --image=gcr.io/$PROJECT_ID/app:$SHORT_SHA
      - --location=us-central1
      - --cluster=my-cluster
      - --output=/workspace/gke-deploy-output

The step runs as UID 65532 rather than root, which the upstream build step does not. Cloud Build points HOME at /builder/home rather than letting the image's own /home/nonroot apply, and run --cluster calls gcloud container clusters get-credentials, which writes SDK state under $HOME. --output also defaults to ./output in the working directory. Pointing HOME and CLOUDSDK_CONFIG at /tmp moves the SDK writes somewhere the nonroot user can always write, and --output is given an explicit path so it does not depend on the working directory. /workspace must itself be writable by UID 65532; this has not been verified against a live Cloud Build run, so confirm it for your project before relying on it.

Configuration

gke-deploy is configured by flags; the environment variables that matter in a container are the Cloud SDK's own:

gcloud is optional. When it is absent from PATH, apply and run fall back to a Go client, and prepare never uses it at all.

Documentation and Resources

• gke-deploy documentation
• Deploying to GKE from Cloud Build
• Recommended Kubernetes labels

What are Chainguard Containers?

Chainguard's free tier of Starter container images are built with Wolfi, our minimal Linux undistro.

All other Chainguard Containers are built with Chainguard OS, Chainguard's minimal Linux operating system designed to produce container images that meet the requirements of a more secure software supply chain.

The main features of Chainguard Containers include:

• Minimal design, without unnecessary software bloat
• Daily builds to ensure container images are up-to-date with available security patches
• High quality build-time SBOMs attesting to the provenance of all artifacts within the image
• Verifiable signatures provided by Sigstore
• Reproducible builds with Cosign and apko (read more about reproducibility)

For cases where you need container images with shells and package managers to build or debug, most Chainguard Containers come paired with a development, or -dev, variant.

In all other cases, including Chainguard Containers tagged as :latest or with a specific version number, the container images include only an open-source application and its runtime dependencies. These minimal container images typically do not contain a shell or package manager.

Although the -dev container image variants have similar security features as their more minimal versions, they include additional software that is typically not necessary in production environments. We recommend using multi-stage builds to copy artifacts from the -dev variant into a more minimal production image.

Need additional packages?

To improve security, Chainguard Containers include only essential dependencies. Need more packages? Chainguard customers can use Custom Assembly to add packages, either through the Console, chainctl, or API.

To use Custom Assembly in the Chainguard Console: navigate to the image you'd like to customize in your Organization's list of images, and click on the Customize image button at the top of the page.

Learn More

Refer to our Chainguard Containers documentation on Chainguard Academy. Chainguard also offers VMs and Libraries — contact us for access.

Trademarks

This software listing is packaged by Chainguard. The trademarks set forth in this offering are owned by their respective companies, and use of them does not imply any affiliation, sponsorship, or endorsement by such companies.