1.12 cloudprovider AAD flake

[jackfrancis]

Saw this on one of three master in an E2E run against 1.12.0:

Oct 04 19:48:57 k8s-master-19753088-0 kubelet[10633]: F1004 19:48:57.959195   10633 server.go:262] failed to run Kubelet: could not init cloud provider "azure": No credentials provided for AAD application
Oct 04 19:48:57 k8s-master-19753088-0 systemd[1]: kubelet.service: Main process exited, code=exited, status=255/n/a
Oct 04 19:48:57 k8s-master-19753088-0 systemd[1]: kubelet.service: Unit entered failed state.
Oct 04 19:48:57 k8s-master-19753088-0 systemd[1]: kubelet.service: Failed with result 'exit-code'.
Oct 04 19:48:58 k8s-master-19753088-0 systemd[1]: kubelet.service: Service hold-off time over, scheduling restart.
Oct 04 19:48:58 k8s-master-19753088-0 systemd[1]: Stopped Kubelet.
Oct 04 19:48:58 k8s-master-19753088-0 systemd[1]: kubelet.service: Start request repeated too quickly.
Oct 04 19:48:58 k8s-master-19753088-0 systemd[1]: Failed to start Kubelet.

[jackfrancis]

$ ls -la /etc/kubernetes/azure.json 
-rw------- 1 root root 0 Oct  4 19:48 /etc/kubernetes/azure.json

[tariq1890]

Could this be related to #3919 ?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contribution. Note that acs-engine is deprecated--see https://github.com/Azure/aks-engine instead.