Nightly testing of integration points (AWS)

[Racer159]

Describe what should be investigated or refactored

Zarf integrates with many external services and each of these has its unique *ilities that are likely to change over time. We should be checking against the following:

• A wide range of kubernetes cluster versions and distros
• A wide range of git repositories (for flux deployments)
• A wide range of image registries (for deployments and zarf package publishing/deploying)

Links to any relevant code

https://github.com/defenseunicorns/zarf/tree/main/.github/workflows

Additional context

This is something that we have discussed for a while, but would require running as a nightly (or other interval) because it will need special accounts that are not easily parallelized and will take a long time to run

[Noxsios]

publishing + pulling from:

• ghcr
• ecr
• distribution/distribution
• dockerhub
• gar
• acr

[Racer159]

For this first go we should probably seek Amazon services and then just setup the framework for adding more. For k8s clusters we should support:

EKS, AKS, OpenShift (https://www.ibm.com/cloud/openshift), and (maybe), GCP

For git we should test:

Gitlab (the only external git provider that supports push to create)

[Racer159]

This may be a good case to better test Windows as well (ref #1513) so that we can build out a matrix for support across OSes, Clusters, Registries and Git Servers.

[Racer159]

This issue will focus on AWS first - specifically - EKS for cluster deployment and ECR for package publish - ECR for package deployment will come later - #1594

[Racer159]

AWS nightlies passed, closing