Addons Host upgrade and Dashboard Container Creating After a fresh install

[AMO-MTSY]

pharos version 1.3.0 cluster.yml host-upgrades: enabled: true schedule: "30 6 *" schedule_window: 1h reboot: true kubernetes-dashboard: enabled: true kubectl get pods --namespace kube-system

• cert-manager-5c7f844c7f-w49n4 2/2 Running 0 32m
• coredns-69c75956f7-g4qx9 1/1 Running 0 33m
• coredns-69c75956f7-gch6s 1/1 Running 0 33m
• etcd-vps504991 1/1 Running 0 32m
• heapster-79b95f7d45-7gn48 1/1 Running 0 32m
• =>host-upgrades-cvkkl 0/1 ContainerCreating 0 32m
• host-upgrades-pkwj7 1/1 Running 0 32m
• kube-apiserver-vps504991 1/1 Running 0 29m
• kube-controller-manager-vps504991 1/1 Running 0 29m
• kube-proxy-n29t7 1/1 Running 0 33m
• kube-proxy-rcnbc 1/1 Running 0 32m
• kube-scheduler-vps504991 1/1 Running 0 31m
• =>kubernetes-dashboard-866596645-dnmkx 0/1 ContainerCreating 0 32m
• metrics-server-57b998f5fc-hsndq 1/1 Running 0 33m
• pharos-proxy-vps573747 1/1 Running 0 32m
• weave-net-ckg6c 2/2 Running 0 33m
• weave-net-rjl4n 2/2 Running 1 32m kubectl get events --all-namespaces | grep upgrade Failed create pod sandbox: rpc error: code = Unknown desc = failed to create pod network sandbox k8s_host-upgrades-cvkkl_kube-system_3d81eaff-ac88-11e8-bb47-fa163e5c7646_0(491704e8232524c84ca66545908c1a852673fe639065ad9f04e969f735041267): failed to find plugin "weave-net" in path [/opt/cni/bin/ /opt/weave-net/bin] kubectl get events --all-namespaces | grep dashboard
• creating: pods "kubernetes-dashboard-866596645-" is forbidden: error looking up service account kube-system/kubernetes-dashboard: serviceaccount "kubernetes-dashboard" not found
• Failed create pod sandbox: rpc error: code = Unknown desc = failed to create pod network sandbox k8s_kubernetes-dashboard-866596645-dnmkx_kube-system_4291ca1f-ac88-11e8-bb47-fa163e5c7646_0(2d7251fcf85275b84ba776977d1d684ad9f6edd5e217b31eb6445f2914afd59c): failed to find plugin "weave-net" in path [/opt/cni/bin/ /opt/weave-net/bin]
• Failed create pod sandbox: rpc error: code = Unknown desc = pod sandbox with name "k8s_kubernetes-dashboard-866596645-dnmkx_kube-system_4291ca1f-ac88-11e8-bb47-fa163e5c7646_0" already exists

[AMO-MTSY]

I think i was using incompatible configs in my cluster.yml: think the example provided in the doc is more suitable to a bare metal setup.

[jnummelin]

Could you show the full cluster.yml? If the problem is common, Pharos should try to protect you by erroring out early instead of setting up the cluster wrongly.

[AMO-MTSY]

Hi, I started with the example cluster.yml provided in the doc. Those are the changes i made to make it work for me: network: provider: calico service_cidr: 172.31.0.0/16 pod_network_cidr: 172.32.0.0/16

Removed "authentication", "audit" and "kube_proxy"