Installing in an air-gapped environment, where there is no internet access available, requires some extra preparation steps before StackState can be installed.
Note that step 2 requires a Docker registry that's available to the Kubernetes cluster and can store all StackState docker images.
Configure Helm
Configure Helm on your local machine to be able to pull the StackState Helm chart.
Make sure to first run helm repo update again to have the latest version of the Helm chart available.
Download the latest StackState helm chart like this:
helm pull stackstate/stackstate-k8s
This results in a file like this stackstate-k8s-1.0.4.tgz. Copy this file (using scp, sftp or any other tool available) to the system from which StackState will be installed.
Copy the StackState docker images
Make sure to first run helm repo update again to have the latest version of the Helm chart available.
chmod +x copy_images.sh
The script can copy images directly from StackState's Quay.io registry to your internal registry. If the internal registry isn't accessible from a computer that has direct internet access an intermediate step is needed.
To copy the images directly from the StackState registry to the internal registry run the script like this to copy the images to the registry at registry.acme.com:5000:
The environment variables are used to setup authentication. If the destination registry doesn't require authentication the DST_REGISTRY_* variables can be omitted.
The script extracts all images from the Helm chart and copies the images to the local registry. Depending on the speed of the internet connection this might take a while.
When it's impossible to directly copy the images to the internal registry the images can be listed using the copy_images.sh script. The best way to download, re-tag, and copy the images to the internal registry depends on the exact circumstances.
Here is an example way of working that uses the copy_images.sh script to produce a list of images and then uses bash scripting to download, re-tag and upload all images. Depending on the exact situation this may need be adapted.
# Produce a list of all StackState images in the stackstate_images.txt file
STS_REGISTRY_USERNAME=noop STS_REGISTRY_PASSWORD=noop ./copy_images.sh -t -d noop | cut -d' ' -f2 > stackstate_images.txt
# Authenticate to the StackState quay.io repositories using the credentials provided by StackState
docker login quay.io
# Save all images to the local file system
mkdir images
while read image; do
name=$(echo "$image" | cut -d'/' -f3)
docker pull --platform linux/amd64 "$image"
docker save "$image" -o "images/${name}.tar"
done < stackstate_images.txt
# Now copy images to the air-gapped environment, for example using scp or sftp. Also copy the stackstate_images.txt file
# On a computer inside the air-gapped environment load, re-tag and push the images, this uses registry.acme.com:5000 as the internal registry
while read image; do
name=$(echo "$image" | cut -d'/' -f3)
target_image="registry.acme.com:5000/stackstate/$name"
docker load -i "images/${name}.tar"
docker tag "$image" "$target_image"
docker push "$target_image"
done < stackstate_images.txt
Prepare local Docker registry configuration
In preparation for the installation of StackState create a local-docker-registry.yaml values file that will be used during the Helm installation of StackState. Include the following configuration in that file, replacing the registry.acme.com:5000 with your internal docker registry where the docker images have been uploaded in the ealier steps.
