update manifest generation to match centralized cloud provider configuration

[yastij]

/kind bug

What steps did you take and what happened: after #471 merged we moved from cloud provider configuration being specified on the machine config to cluster configuration.

Currently

docker run --rm   -v "$(pwd)":/out   -v "$(pwd)/envvars.txt":/envvars.txt:ro   gcr.io/cluster-api-provider-vsphere/release/manifests:latest   -c management-cluster

does not reflect that and cluster bootstrapping fails with:

W0812 14:38:33.977694       1 machine_controller.go:228] Failed to create machine "management-cluster-controlplane-1": unable to get datastore for "default/management-cluster/management-cluster-controlplane-1": default datastore resolves to multiple instances, please specify

What did you expect to happen:
bootstrapping the management cluster

Anything else you would like to add:

/assign

Environment:

• Cluster-api-provider-vsphere version:
• Kubernetes version: (use kubectl version):
• OS (e.g. from /etc/os-release):

[yastij]

@akutz - seems like on release/manifests:latest templates under /build and /out/cmd are diverging

[akutz]

/assign @codenrhoden

[akutz]

Hi @codenrhoden,

Do you need help looking at this? I also wonder if it's still an issue in CAPV v1alpha2.

[codenrhoden]

I can definitely look at this. With E2E tests working again, it indeed may no longer be a problem. I need to come up to speed on the changes over the last couple of weeks, but I will do so.

[yastij]

I'm no longer able to reproduce it

/close

[k8s-ci-robot]

@yastij: Closing this issue.

In response to this:

I'm no longer able to reproduce it

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.