Kubernetes 工作节点中的网络服务

Question

我已经按照https://kubernetes.io/docs/setup/independent/create-cluster-kubeadm/ 安装了 3 台服务器 kubernetes 设置

我在主节点创建了calico网络服务。我的问题是我也应该在工作节点中创建印花布服务吗？

当我创建 pod 时，我在工作节点中遇到错误

ngwhq_kube-system(e17770a3-8507-11e8-962c-0ac29e406ef0)"
Jul 11 13:25:05 ip-172-31-20-212 kubelet: I0711 13:25:05.144142   23325 kuberuntime_manager.go:767] Back-off 5m0s restarting failed container=calico-node pod=calico-node-ngwhq_kube-system(e17770a3-8507-11e8-962c-0ac29e406ef0)
Jul 11 13:25:05 ip-172-31-20-212 kubelet: E0711 13:25:05.144169   23325 pod_workers.go:186] Error syncing pod e17770a3-8507-11e8-962c-0ac29e406ef0 ("calico-node-ngwhq_kube-system(e17770a3-8507-11e8-962c-0ac29e406ef0)"), skipping: failed to "StartContainer" for "calico-node" with CrashLoopBackOff: "Back-off 5m0s restarting failed container=calico-node pod=calico-node-ngwhq_kube-system(e17770a3-8507-11e8-962c-0ac29e406ef0)"
Jul 11 13:25:07 ip-172-31-20-212 kubelet: E0711 13:25:07.221953   23325 cni.go:280] Error deleting network: context deadline exceeded
Jul 11 13:25:07 ip-172-31-20-212 kubelet: E0711 13:25:07.222595   23325 remote_runtime.go:115] StopPodSandbox "22fe8b5db360011aa79afadfe91a46bfef0322092478d378ef657d3babfc1326" from runtime service failed: rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod "test2-597bdc85dc-k2xsm_default" network: context deadline exceeded
Jul 11 13:25:07 ip-172-31-20-212 kubelet: E0711 13:25:07.222630   23325 kuberuntime_manager.go:799] Failed to stop sandbox {"docker" "22fe8b5db360011aa79afadfe91a46bfef0322092478d378ef657d3babfc1326"}
Jul 11 13:25:07 ip-172-31-20-212 kubelet: E0711 13:25:07.222664   23325 kuberuntime_manager.go:594] killPodWithSyncResult failed: failed to "KillPodSandbox" for "67e18616-850d-11e8-962c-0ac29e406ef0" with KillPodSandboxError: "rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod \"test2-597bdc85dc-k2xsm_default\" network: context deadline exceeded"
Jul 11 13:25:07 ip-172-31-20-212 kubelet: E0711 13:25:07.222685   23325 pod_workers.go:186] Error syncing pod 67e18616-850d-11e8-962c-0ac29e406ef0 ("test2-597bdc85dc-k2xsm_default(67e18616-850d-11e8-962c-0ac29e406ef0)"), skipping: failed to "KillPodSandbox" for "67e18616-850d-11e8-962c-0ac29e406ef0" with KillPodSandboxError: "rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod \"test2-597bdc85dc-k2xsm_default\" network: context deadline exceeded"
Jul 11 13:25:12 ip-172-31-20-212 kubelet: E0711 13:25:12.007944   23325 cni.go:280] Error deleting network: context deadline exceeded
Jul 11 13:25:12 ip-172-31-20-212 kubelet: E0711 13:25:12.008783   23325 remote_runtime.go:115] StopPodSandbox "4b14d68c7bc892594dedd1f62d92414574a3fb00873a805b62707c7a63bfdfe7" from runtime service failed: rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod "test2-597bdc85dc-qmc85_default" network: context deadline exceeded
Jul 11 13:25:12 ip-172-31-20-212 kubelet: E0711 13:25:12.008819   23325 kuberuntime_gc.go:153] Failed to stop sandbox "4b14d68c7bc892594dedd1f62d92414574a3fb00873a805b62707c7a63bfdfe7" before removing: rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod "test2-597bdc85dc-qmc85_default" network: context deadline exceeded
Jul 11 13:25:19 ip-172-31-20-212 kubelet: W0711 13:25:19.145386   23325 cni.go:243] CNI failed to retrieve network namespace path: cannot find network namespace for the terminated container "22fe8b5db360011aa79afadfe91a46bfef0322092478d378ef657d3babfc1326"

我也尝试使用下面提到的命令在工作节点中安装 calico 网络，但运气不佳..

kubectl apply -f  https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml



unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused
unable to recognize "https://docs.projectcalico.org/v3.1/getting-started/kubernetes/installation/hosted/kubeadm/1.7/calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp 127.0.0.1:8080: connect: connection refused

Answer 1

每个节点都需要运行calico服务，这是常识。