Include all APIserver addresses for nodeup config

[rifelpet]

With this plus #16812 i'm able to get an ipv6 dns=none cluster to pass validation.

Before we were only including ipv4 addresses because those are the only CIDRs included in the cluster spec - ipv6 CIDRs are provided by AWS. This resulted in nodes failing to bootstrap because they couldn't reach the kops-controller endpoint:

Sep 06 01:23:11 i-0250853b64092f19d nodeup[1323]: W0906 01:23:11.157536 1323 main.go:133] got error running nodeup (will retry in 30s): failed to get node config from server: Post "https://172.20.5.248:3988/bootstrap": dial tcp 172.20.5.248:3988: connect: network is unreachable

Even though the ipv6 address and load balancer DNS name both work:

$ curl -6 -k  'https://[2600:1f14:1800:ab02:ba1a:4b60:b424:d3f3]:3988/bootstrap'
failed to verify token

curl -k https://api-peter-ipv6-k8s--k283sk-ef9b199ef8b93ba4.elb.us-west-2.amazonaws.com:3988/bootstrap
failed to verify token

With this change, the nodeup config userdata changes as such:

Will modify resources:
  LaunchTemplate/control-plane-us-west-2a.masters.peter-rifel-ipv6.k8s.local
  	UserData
  	                    	...
  	                    	  APIServerIPs:
  	                    	  - 172.20.5.248
  	                    	+ - 2600:1f14:1800:ab02:ba1a:4b60:b424:d3f3
  	                    	+ - api-peter-ipv6-k8s--k283sk-ef9b199ef8b93ba4.elb.us-west-2.amazonaws.com
  	                    	  CloudProvider: aws
  	                    	  ClusterName: peter-ipv6.k8s.local
  	                    	...


  LaunchTemplate/nodes-us-west-2a.peter-rifel-ipv6.k8s.local
  	UserData
  	                    	...
  	                    	  APIServerIPs:
  	                    	  - 172.20.5.248
  	                    	+ - 2600:1f14:1800:ab02:ba1a:4b60:b424:d3f3
  	                    	+ - api-peter-ipv6-k8s--k283sk-ef9b199ef8b93ba4.elb.us-west-2.amazonaws.com
  	                    	  CloudProvider: aws
  	                    	  ClusterName: peter-ipv6.k8s.local
  	                    	...
  	                    	    servers:
  	                    	    - https://172.20.5.248:3988/
  	                    	+   - https://[2600:1f14:1800:ab02:ba1a:4b60:b424:d3f3]:3988/
  	                    	+   - https://api-peter-ipv6-k8s--k283sk-ef9b199ef8b93ba4.elb.us-west-2.amazonaws.com:3988/
  	                    	  InstanceGroupName: nodes-us-west-2a
  	                    	  InstanceGroupRole: Node
  	                    	...


Must specify --yes to apply changes

Marking this as draft because its possible we could find a better approach, and because this may break other cluster configurations

[k8s-ci-robot]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[rifelpet]

From office hours: I will add back the filtering logic but always include any ipv6 addresses. It will still exclude the DNS names though.

[rifelpet]

/test pull-kops-e2e-cni-cilium-ipv6

[rifelpet]

the cluster validated and e2e tests started, but the prow job pod was interrupted

/test pull-kops-e2e-cni-cilium-ipv6

[rifelpet]

tests pass too 🎉
/cc @hakman

[rifelpet]

It looks like this is only needed for dns=none clusters. normal DNS clusters are already passing with my IMDS and controller-runtime changes: https://testgrid.k8s.io/kops-ipv6#kops-aws-cni-calico-ipv6-flatcar

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: hakman

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:

• OWNERS [hakman]

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

[hakman]

Awesome work!