Add VPC claiming for external cluster types

[gabrieljackson]

In an effort to support EKS clusters in our standard production AWS accounts, this change allows for importing clusters with VPC claiming. Clusters imported in this way will be able to perform AWS resource lookups to manage installations with pgbouncer, bifrost, and similar dependencies.

This is example metadata for an external cluster with a claimed VPC:

        "ID": "k5mk7fr1k3r8tmi46dbbf5wj6w",
        "State": "stable",
        "Provider": "external",
        "ProviderMetadataExternal": {
            "HasAWSInfrastructure": true
        },
        "Provisioner": "external",
        "ProvisionerMetadataExternal": {
            "Name": "k5mk7fr1k3r8tmi46dbbf5wj6w-external-k8s",
            "SecretName": "kubeconfig-net-calico",
            "Version": "1.29.4-eks-036c24b",
            "VPC": "vpc-0846s82b132aad7e7",
            "Warnings": [
                "Health Check: failed to find any deployments in namespace bifrost"
            ]
        },

A follow-up PR will be needed to incorporate new VPC lookup filters so that kops clusters can't accidentally claim a VPC with EKS clusters in them.

Fixes https://mattermost.atlassian.net/browse/CLD-7985

Add VPC claiming for external cluster types

[andrleite]

@gabrieljackson nice work on this! One question. So for now, Will we only check for the most important utilities and put a warning in case they don't exist? Bifrost isn't in argo for example, will this warning block cluster from receiving installations?

[nickmisasi]

👍

[gabrieljackson]

One question. So for now, Will we only check for the most important utilities and put a warning in case they don't exist? Bifrost isn't in argo for example, will this warning block cluster from receiving installations?

@andrleite yep, currently the logic is warn and proceed for utilities. We can tweak this in the future if we like as well as add more checks, but we may need to get some testing in on this behavior to see what will work best for us.