flannel pod (k8z_install-cni) giving problems when a new node joins kubernetes cluster (v1.5.1)

1/12/2017

I am trying to join a node to my kubernetes cluster using kubeadm

I followed this tutorial

My kubernetes version is 1.5.1

I got the flannel yaml from this link

I didn't get any problem during installation but wen I join a new node I can see:

[root@kub1 ~]# kubectl get pods --namespace kube-system
NAME                                 READY     STATUS             RESTARTS   AGE
dummy-2088944543-b18d7               1/1       Running            0          19h
etcd-kub1.local                      1/1       Running            0          19h
kube-apiserver-kub1.local            1/1       Running            0          19h
kube-controller-manager-kub1.local   1/1       Running            0          19h
kube-discovery-1769846148-zcxrx      1/1       Running            0          19h
kube-dns-2924299975-zbnjb            4/4       Running            0          19h
kube-flannel-ds-5ns9q                2/2       Running            1          19h
kube-flannel-ds-thssr                1/2       CrashLoopBackOff   13         44m
kube-proxy-9s15v                     1/1       Running            0          19h
kube-proxy-sfnl0                     1/1       Running            0          44m
kube-scheduler-kub1.local            1/1       Running            0          19h

I can see the following issue on the pod description:

<invalid>     <invalid>       2       {kubelet kub2.local}                                    Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "install-cni" with CrashLoopBackOff: "Back-off 20s restarting failed container=install-cni pod=kube-flannel-ds-thssr_kube-system(322354fd-d925-11e6-a4aa-0050568da433)"

And the docker logs from the container k8z_install-cni:

+ cp -f /etc/kube-flannel/cni-conf.json /etc/cni/net.d/10-flannel.conf
+ true
+ sleep 3600

QUESTION: How can I fix this?

thank you very much

-- Manuel Sopena Ballesteros
flannel
kubernetes

0 Answers