CAPI: Do not normalize Node IDs outside of CAPI provider

[JoelSpeed]

When #1866 introduced the CAPI provider, all Provider IDs from nodes were being normalized so that internally, there were no differences (eg some places using normalized, some not)

However, the clusterstateregistry and anything outside of the CPAI provider does not expect the instance IDs to be normalized, and so nodes were being marked as unregistered due to the normalized ID not matching the non-normalized ID coming from the Nodes on the cluster.

This PR makes it so that Instance IDs fetched and passed externally are not normalized, fixing the unregistered node logic in the clusterstateregistry when using the CAPI provider.

/cc @enxebre @elmiko

[JoelSpeed]

/assign @hardikdr

[elmiko]

nice find Joel!

/lgtm

[enxebre]

Referencing the particular commit which normalized to be able to match nodes/machines d9e3197

However that ^^ only partially fixed the problem. This will still break for some scenarios e.g upstream cluster-api-provider-aws as its providerID does not includes the AZ consistently with the control manager cloud provider. We should probably fix there as well -> https://github.com/kubernetes-sigs/cluster-api-provider-aws/blob/22be31dd76da04409c425c77e2967db327f409a4/controllers/awsmachine_controller.go#L406

[JoelSpeed]

/hold

Need to make sure this doesn't merge until @enxebre's comment is resolved

[JoelSpeed]

This will need resolving before this PR can be unblocked kubernetes-sigs/cluster-api-provider-aws#1693

[temujin9]

I can confirm that this patch -- provided as an image by @elmiko -- exhibits the same behavior on AWS without the @enxebre's other change. It looks like the other change just merged; I'll try to test a build with it shortly.

[temujin9]

That second change finished the trick: my autoscaled nodes now persist for as long as load does, rather than rolling every 10 minutes.

[enxebre]

/approve

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: enxebre

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• cluster-autoscaler/cloudprovider/clusterapi/OWNERS [enxebre]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[elmiko]

/lgtm

[JoelSpeed]

/hold cancel