Support creating Clusters using an http proxy in cluster-cloud-director

[cornelius-keller]

Support creating clusters using HTTP_PROXY in cluster-cloud-director.
Please make sure that we are using the same interface (helm variables) as team hydra. See giantswarm/cluster-aws#105

Tasks

• Make sure that app platform still supports running behind a proxy so that we can get things bootstrapped.

[cornelius-keller]

Might already be done, needs to be verified @vxav

[vxav]

We already have values used for nodes. Maybe we can reuse them for the rest.

https://github.com/giantswarm/cluster-cloud-director/blob/main/helm/cluster-cloud-director/values.yaml#L46-L49

[vxav]

PR to align proxy values with CAPA => giantswarm/cluster-cloud-director#30
No change in functionality, we still use the native capvcd integration.
This only set environment variables on the nodes, not the workloads.

I suppose we just need to test this now. @glitchcrab we can have a look at your proxy and deploy a test cluster using it in the values?

[vxav]

The HTTPS Proxy env variable is configured for containerd on the nodes and can pull images fine.
However it doesn't pupolate the HTTP env variable because of a typo. The maintainers are fixing it.

Will require to update controller+cluster chart to stay current.

[vxav]

Closing this as it is done from a cluster chart perspective.

[vxav]

After discussions, we want to use kubeadmconfigspec to make it consistent across provider and not use the CAPVCD built-in implementation with cloud-init.

[gawertm]

@vxav I think this story is already implemented? current proxy implementation by @bavarianbidi was tested on CAPVCD

[bavarianbidi]

The current implementation in cluster-cloud-director based on this PR in cluster-api-provider-cloud-director make it possible to create a cluster behind a proxy.

in upstream issue #299 we track the current process of our forked changes (as in one of the recent meetings with CAPVCD-maintainers we decided with them that the current build-in solution isn't very flexible)