kubeadm init 1.9 hangs with vsphere cloud provider - kubernetes

Question

kubeadm init seems to be hanging when I started using vsphere cloud provider. Followed instructions from here - Anybody got it working with 1.9?
root#master-0:~# kubeadm init --config /tmp/kube.yaml
[init] Using Kubernetes version: v1.9.1
[init] Using Authorization modes: [Node RBAC]
[preflight] Running pre-flight checks.
[WARNING Hostname]: hostname "master-0" could not be reached
[WARNING Hostname]: hostname "master-0" lookup master-0 on 8.8.8.8:53: no such host
[WARNING FileExisting-crictl]: crictl not found in system path
[preflight] Starting the kubelet service
[certificates] Generated ca certificate and key.
[certificates] Generated apiserver certificate and key.
[certificates] apiserver serving cert is signed for DNS names [master-0 kubernetes kubernetes.default kubernetes.default.svc kubernetes.default.svc.cluster.local] and IPs [10.96.0.1 10.11.0.101]
[certificates] Generated apiserver-kubelet-client certificate and key.
[certificates] Generated sa key and public key.
[certificates] Generated front-proxy-ca certificate and key.
[certificates] Generated front-proxy-client certificate and key.
[certificates] Valid certificates and keys now exist in "/etc/kubernetes/pki"
[kubeconfig] Wrote KubeConfig file to disk: "admin.conf"
[kubeconfig] Wrote KubeConfig file to disk: "kubelet.conf"
[kubeconfig] Wrote KubeConfig file to disk: "controller-manager.conf"
[kubeconfig] Wrote KubeConfig file to disk: "scheduler.conf"
[controlplane] Wrote Static Pod manifest for component kube-apiserver to "/etc/kubernetes/manifests/kube-apiserver.yaml"
[controlplane] Wrote Static Pod manifest for component kube-controller-manager to "/etc/kubernetes/manifests/kube-controller-manager.yaml"
[controlplane] Wrote Static Pod manifest for component kube-scheduler to "/etc/kubernetes/manifests/kube-scheduler.yaml"
[init] Waiting for the kubelet to boot up the control plane as Static Pods from directory "/etc/kubernetes/manifests".
[init] This might take a minute or longer if the control plane images have to be pulled.
Master os details
root#master-0:~# uname -r
4.4.0-21-generic
root#master-0:~# docker version
Client:
Version: 17.03.2-ce
API version: 1.27
Go version: go1.7.5
Git commit: f5ec1e2
Built: Tue Jun 27 03:35:14 2017
OS/Arch: linux/amd64
Server:
Version: 17.03.2-ce
API version: 1.27 (minimum version 1.12)
Go version: go1.7.5
Git commit: f5ec1e2
Built: Tue Jun 27 03:35:14 2017
OS/Arch: linux/amd64
Experimental: false
root#master-0:~# cat /etc/os-release
NAME="Ubuntu"
VERSION="16.04 LTS (Xenial Xerus)"
ID=ubuntu
kubelet service seems to be running fine
root#master-0:~# systemctl status kubelet
● kubelet.service - kubelet: The Kubernetes Node Agent
Loaded: loaded (/lib/systemd/system/kubelet.service; enabled; vendor preset: enabled)
Drop-In: /etc/systemd/system/kubelet.service.d
└─10-kubeadm.conf, 90-local-extras.conf
Active: active (running) since Mon 2018-01-22 11:25:00 UTC; 13min ago
Docs: http://kubernetes.io/docs/
Main PID: 4270 (kubelet)
Tasks: 13 (limit: 512)
Memory: 37.6M
CPU: 11.626s
CGroup: /system.slice/kubelet.service
└─4270 /usr/bin/kubelet --bootstrap-kubeconfig=/etc/kubernetes/bootstrap-kubelet.conf --kubeco
nfig=/etc/kubernetes/kubelet.conf --pod-manifest-path=/etc/kubernetes/manifests --allow-privileged=true
--network-plugin=cni --cni-conf-dir=/etc/cni/net.d --cni-bin-dir=/opt/cni/bin --cluster-dns=10.96.0.10
--cluster-domain=cluster.local --authorization-mode=Webhook --client-ca-file=/etc/kubernetes/pki/ca.cr
t --cadvisor-port=0 --rotate-certificates=true --cert-dir=/var/lib/kubelet/pki
journalctl -f -u kubelet has some connection refused errors which probably networking service is missing. Those errors should go away when networking service is installed post kubeadm init
Jan 22 11:17:45 localhost kubelet[1184]: I0122 11:17:45.759764 1184 feature_gate.go:220] feature gat
es: &{{} map[]}
Jan 22 11:17:45 localhost kubelet[1184]: I0122 11:17:45.761350 1184 controller.go:114] kubelet confi
g controller: starting controller
Jan 22 11:17:45 localhost kubelet[1184]: I0122 11:17:45.762632 1184 controller.go:118] kubelet confi
g controller: validating combination of defaults and flags
Jan 22 11:17:46 localhost systemd[1]: Started Kubernetes systemd probe.
Jan 22 11:17:46 localhost kubelet[1184]: W0122 11:17:46.070619 1184 cni.go:171] Unable to update cni
config: No networks found in /etc/cni/net.d
Jan 22 11:17:46 localhost kubelet[1184]: I0122 11:17:46.081384 1184 server.go:182] Version: v1.9.1
Jan 22 11:17:46 localhost kubelet[1184]: I0122 11:17:46.081417 1184 feature_gate.go:220] feature gat
es: &{{} map[]}
Jan 22 11:17:46 localhost kubelet[1184]: I0122 11:17:46.082271 1184 plugins.go:101] No cloud provide
r specified.
Jan 22 11:17:46 localhost kubelet[1184]: error: failed to run Kubelet: unable to load bootstrap kubecon
fig: stat /etc/kubernetes/bootstrap-kubelet.conf: no such file or directory
Jan 22 11:17:46 localhost systemd[1]: kubelet.service: Main process exited, code=exited, status=1/FAILU
RE
Jan 22 11:17:46 localhost systemd[1]: kubelet.service: Unit entered failed state.
Jan 22 11:17:46 localhost systemd[1]: kubelet.service: Failed with result 'exit-code'.
Jan 22 11:17:56 localhost systemd[1]: kubelet.service: Service hold-off time over, scheduling restart.
Jan 22 11:17:56 localhost systemd[1]: Stopped kubelet: The Kubernetes Node Agent.
Jan 22 11:17:56 localhost systemd[1]: Started kubelet: The Kubernetes Node Agent.
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.410883 1229 feature_gate.go:220] feature gat
es: &{{} map[]}
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.411198 1229 controller.go:114] kubelet confi
g controller: starting controller
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.411353 1229 controller.go:118] kubelet confi
g controller: validating combination of defaults and flags
Jan 22 11:17:56 localhost systemd[1]: Started Kubernetes systemd probe.
Jan 22 11:17:56 localhost kubelet[1229]: W0122 11:17:56.424264 1229 cni.go:171] Unable to update cni
config: No networks found in /etc/cni/net.d
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.429102 1229 server.go:182] Version: v1.9.1
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.429156 1229 feature_gate.go:220] feature gat
es: &{{} map[]}
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.429247 1229 plugins.go:101] No cloud provide
r specified.
Jan 22 11:17:56 localhost kubelet[1229]: E0122 11:17:56.461608 1229 certificate_manager.go:314] Fail
ed while requesting a signed certificate from the master: cannot create certificate signing request: Po
st https://10.11.0.101:6443/apis/certificates.k8s.io/v1beta1/certificatesigningrequests: dial tcp 10.11
.0.101:6443: getsockopt: connection refused
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.491374 1229 server.go:428] --cgroups-per-qos
enabled, but --cgroup-root was not specified. defaulting to /
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.492069 1229 container_manager_linux.go:242]
container manager verified user specified cgroup-root exists: /
Jan 22 11:17:56 localhost kubelet[1229]: I0122 11:17:56.492102 1229 container_manager_linux.go:247]
Creating Container Manager object based on Node Config: {RuntimeCgroupsName: SystemCgroupsName: Kubelet
CgroupsName: ContainerRuntime:docker CgroupsPerQOS:true CgroupRoot:/ CgroupDriver:cgroupfs KubeletRootD
ir:/var/lib/kubelet ProtectKernelDefaults:false NodeAllocatableConfig:{KubeReservedCgroupName: SystemRe
servedCgroupName: EnforceNodeAllocatable:map[pods:{}] KubeReserved:map[] SystemReserved:map[] HardEvict
ionThresholds:[{Signal:memory.available Operator:LessThan Value:{Quantity:100Mi Percentage:0} GracePeri
od:0s MinReclaim:<nil>} {Signal:nodefs.available Operator:LessThan Value:{Quantity:<nil> Percentage:0.1
} GracePeriod:0s MinReclaim:<nil>} {Signal:nodefs.inodesFree Operator:LessThan Value:{Quantity:<nil> Pe
rcentage:0.05} GracePeriod:0s MinReclaim:<nil>} {Signal:imagefs.available Operator:LessThan Value:{Quan
tity:<nil> Percentage:0.15} GracePeriod:0s MinReclaim:<nil>}]} ExperimentalQOSReserved:map[] Experiment
alCPUManagerPolicy:none ExperimentalCPUManagerReconcilePeriod:10s}
docker ps, controller & scheduler logs
root#master-0:~# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
6db549891439 677911f7ae8f "kube-scheduler --..." About an hour ago Up About an hour k8s_kube-scheduler_kube-scheduler-master-0_kube-system_df32e281019039e73be77e3f53d09596_0
4f7ddefbd86e 4978f9a64966 "kube-controller-m..." About an hour ago Up About an hour k8s_kube-controller-manager_kube-controller-manager-master-0_kube-system_34bad395be69e74db6304d6c4218c536_0
18604db89db6 gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD_kube-scheduler-master-0_kube-system_df32e281019039e73be77e3f53d09596_0
252b86ea4b5e gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD_kube-controller-manager-master-0_kube-system_34bad395be69e74db6304d6c4218c536_0
4021061bf8a6 gcr.io/google_containers/pause-amd64:3.0 "/pause" About an hour ago Up About an hour k8s_POD_kube-apiserver-master-0_kube-system_7a3ae9279d0ca7b4ada8333fbe7442ed_0
4f94163d313b gcr.io/google_containers/etcd-amd64:3.1.10 "etcd --name=etcd0..." About an hour ago Up About an hour 0.0.0.0:2379-2380->2379-2380/tcp, 0.0.0.0:4001->4001/tcp, 7001/tcp etcd
root#master-0:~# docker logs -f 4f7ddefbd86e
I0122 11:25:06.253706 1 controllermanager.go:108] Version: v1.9.1
I0122 11:25:06.258712 1 leaderelection.go:174] attempting to acquire leader lease...
E0122 11:25:06.259448 1 leaderelection.go:224] error retrieving resource lock kube-system/kube-controller-manager: Get https://10.11.0.101:6443/api/v1/namespaces/kube-system/endpoints/kube-controller-manager: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:09.711377 1 leaderelection.go:224] error retrieving resource lock kube-system/kube-controller-manager: Get https://10.11.0.101:6443/api/v1/namespaces/kube-system/endpoints/kube-controller-manager: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:13.969270 1 leaderelection.go:224] error retrieving resource lock kube-system/kube-controller-manager: Get https://10.11.0.101:6443/api/v1/namespaces/kube-system/endpoints/kube-controller-manager: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:17.564964 1 leaderelection.go:224] error retrieving resource lock kube-system/kube-controller-manager: Get https://10.11.0.101:6443/api/v1/namespaces/kube-system/endpoints/kube-controller-manager: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:20.616174 1 leaderelection.go:224] error retrieving resource lock kube-system/kube-controller-manager: Get https://10.11.0.101:6443/api/v1/namespaces/kube-system/endpoints/kube-controller-manager: dial tcp 10.11.0.101:6443: getsockopt: connection refused
root#master-0:~# docker logs -f 6db549891439
W0122 11:25:06.285765 1 server.go:159] WARNING: all flags than --config are deprecated. Please begin using a config file ASAP.
I0122 11:25:06.292865 1 server.go:551] Version: v1.9.1
I0122 11:25:06.295776 1 server.go:570] starting healthz server on 127.0.0.1:10251
E0122 11:25:06.295947 1 reflector.go:205] k8s.io/kubernetes/vendor/k8s.io/client-go/informers/factory.go:86: Failed to list *v1beta1.ReplicaSet: Get https://10.11.0.101:6443/apis/extensions/v1beta1/replicasets?limit=500&resourceVersion=0: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:06.296027 1 reflector.go:205] k8s.io/kubernetes/vendor/k8s.io/client-go/informers/factory.go:86: Failed to list *v1.ReplicationController: Get https://10.11.0.101:6443/api/v1/replicationcontrollers?limit=500&resourceVersion=0: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:06.296092 1 reflector.go:205] k8s.io/kubernetes/vendor/k8s.io/client-go/informers/factory.go:86: Failed to list *v1.Service: Get https://10.11.0.101:6443/api/v1/services?limit=500&resourceVersion=0: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:06.296160 1 reflector.go:205] k8s.io/kubernetes/plugin/cmd/kube-scheduler/app/server.go:590: Failed to list *v1.Pod: Get https://10.11.0.101:6443/api/v1/pods?fieldSelector=spec.schedulerName%3Ddefault-scheduler%2Cstatus.phase%21%3DFailed%2Cstatus.phase%21%3DSucceeded&limit=500&resourceVersion=0: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:06.296218 1 reflector.go:205] k8s.io/kubernetes/vendor/k8s.io/client-go/informers/factory.go:86: Failed to list *v1beta1.StatefulSet: Get https://10.11.0.101:6443/apis/apps/v1beta1/statefulsets?limit=500&resourceVersion=0: dial tcp 10.11.0.101:6443: getsockopt: connection refused
E0122 11:25:06.297374 1 reflector.go:205] k8s.io/kubernetes/vendor/k8s.io/client-go/informers/factory.go:86: Failed to list *v1.PersistentVolume: Get https://10.11.0.101:6443/api/v1/persistentvolumes?limit=500&resourceVersion=0: dial tcp 10.11.0.101:6443: getsockopt: connection refused

Answer 1

There was a bug in the controller manager when starting with the vsphere cloud provider. See https://github.com/kubernetes/kubernetes/issues/57279, fixed in 1.9.2