kube-proxy migration tests are failing: "metrics/v1alpha1: the server could not find the requested resource"

[MrHohn]

Is this a BUG REPORT or FEATURE REQUEST?:

kind bug

What happened:
https://k8s-testgrid.appspot.com/sig-network#gci-gce-latest-upgrade-kube-proxy-ds
&
https://k8s-testgrid.appspot.com/sig-network#gci-gce-latest-downgrade-kube-proxy-ds
are failing consistently since 9/15 with below error message:

Sep 19 13:42:51.893: Couldn't delete ns: "e2e-tests-sig-network-kube-proxy-upgrade-vjjgq": unable to retrieve the complete list of server APIs: metrics/v1alpha1: the server could not find the requested resource (&discovery.ErrGroupDiscoveryFailed{Groups:map[schema.GroupVersion]error{schema.GroupVersion{Group:"metrics", Version:"v1alpha1"}:(*errors.StatusError)(0xc4211e82d0)}})

Example failed build: https://k8s-gubernator.appspot.com/build/kubernetes-jenkins/logs/ci-kubernetes-e2e-gci-gce-latest-upgrade-kube-proxy-ds/1018

Checked the commits diff, seems likely to be caused by #50012.

@deads2k @k8s-mirror-api-machinery-bugs

Anything else we need to know?:

To make thing clear, these two test suites:

1. Upgrade cluster from release/1.8 to master. Or
2. Upgrade cluster from release/1.8 to master and then downgrade cluster to release/1.8 again.

[MrHohn]

/kind bug

[k8s-github-robot]

@MrHohn
There are no sig labels on this issue. Please add a sig label by:

1. mentioning a sig: @kubernetes/sig-<group-name>-<group-suffix>
   e.g., @kubernetes/sig-contributor-experience-<group-suffix> to notify the contributor experience sig, OR

2. specifying the label manually: /sig <label>
   e.g., /sig scalability to apply the sig/scalability label

Note: Method 1 will trigger an email to the group. You can find the group list here and label list here.
The <group-suffix> in the method 1 has to be replaced with one of these: bugs, feature-requests, pr-reviews, test-failures, proposals

[MrHohn]

/sig api-machinery

[mbohlool]

cc @derekwaynecarr

[MrHohn]

Test is no longer failing since 09/22/2017.