kubernetes: Latest alpha fails to run flannel plugin on master node

Is this a BUG REPORT or FEATURE REQUEST? (choose one): BUG REPORT

Kubernetes version (use kubectl version):

Client Version: version.Info{Major:"1", Minor:"6+", GitVersion:"v1.6.0-alpha.3", GitCommit:"5802799e56c7fcd1638e5848a13c5f3b0b1479ab", GitTreeState:"clean", BuildDate:"2017-02-16T19:27:36Z", GoVersion:"go1.7.4", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"6+", GitVersion:"v1.6.0-alpha.3", GitCommit:"5802799e56c7fcd1638e5848a13c5f3b0b1479ab", GitTreeState:"clean", BuildDate:"2017-02-16T19:17:01Z", GoVersion:"go1.7.4", Compiler:"gc", Platform:"linux/amd64"}

Environment:

• Cloud provider or hardware configuration: Occurs on bare metal and AWS.
• OS (e.g. from /etc/os-release):

NAME="CentOS Linux"
VERSION="7 (Core)"
ID="centos"
ID_LIKE="rhel fedora"
VERSION_ID="7"
PRETTY_NAME="CentOS Linux 7 (Core)"
ANSI_COLOR="0;31"
CPE_NAME="cpe:/o:centos:centos:7"
HOME_URL="https://www.centos.org/"
BUG_REPORT_URL="https://bugs.centos.org/"

CENTOS_MANTISBT_PROJECT="CentOS-7"
CENTOS_MANTISBT_PROJECT_VERSION="7"
REDHAT_SUPPORT_PRODUCT="centos"
REDHAT_SUPPORT_PRODUCT_VERSION="7"

• Kernel (e.g. uname -a): Linux k8s-master 3.10.0-514.6.1.el7.x86_64 #1 SMP Wed Jan 18 13:06:36 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux
• Install tools: kubeadm
• Others:

What happened: Installing the Flannel plugin via

kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel-rbac.yml
kubectl create clusterrolebinding flannel --clusterrole=flannel --serviceaccount=kube-system:flannel
kubectl --namespace kube-system apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml

and the flannel pod is never started on the master node. This seems to cause the kube-dns pod to fail with:

E0219 20:45:48.332818       1 reflector.go:199] k8s.io/dns/vendor/k8s.io/client-go/tools/cache/reflector.go:94: Failed to list *v1.Endpoints: Get https://10.96.0.1:443/api/v1/endpoints?resourceVersion=0: dial tcp 10.96.0.1:443: connect: network is unreachable

The flannel pod starts fine on workers. To get DNS working I delete the kube-dns pod:

kubectl delete pod $(kubectl get pod --selector=k8s-app=kube-dns -n kube-system -o jsonpath='{.items[*].metadata.name}') -n kube-system

which seems to get DNS working, but flannel pod is still not running on the master node.

What you expected to happen: Flannel should be running on the master nodes, and the path /etc/cni/net.d/10-flannel.conf should be mounted on the host as specified in https://github.com/coreos/flannel/blob/master/Documentation/kube-flannel.yml.

How to reproduce it (as minimally and precisely as possible): Initialize the cluster via kubeadm, and install flannel plugin

Anything else we need to know:

$ kubelet --version
Kubernetes v1.6.0-alpha.3

$ kubeadm version
kubeadm version: version.Info{Major:"1", Minor:"6+", GitVersion:"v1.6.0-alpha.3", GitCommit:"5802799e56c7fcd1638e5848a13c5f3b0b1479ab", GitTreeState:"clean", BuildDate:"2017-02-16T19:17:01Z", GoVersion:"go1.7.4", Compiler:"gc", Platform:"linux/amd64"}

@luxas