flannel: k8s 1.19.0 with kube-flannel 0.12 Error registering network: failed to acquire lease: node "nodeName" pod cidr not assigned

Running a kubectl patch seems to do the job:

kubectl patch node $(hostname) -p '{"spec":{"podCIDR":"10.100.0.1/24"}}'
kubectl delete -f pod-network-flannel.yml
kubectl apply -f pod-network-flannel.yml

But why do I have to manually patch my node for flannel to work? Isn’t it supposed to automagically get the pod subnet from the control plane?

@kele1997 Thank you for the quick reply. I tried what you said, now I get:

can not mix '--config' with arguments [pod-network-cidr]
To see the stack trace of this error execute with --v=5 or higher

kubectl logs pod/kube-flannel-ds-x5qgg -n kube-system I1204 11:18:34.968447 1 main.go:217] CLI flags config: {etcdEndpoints:http://127.0.0.1:4001,http://127.0.0.1:2379 etcdPrefix:/coreos.com/network etcdKeyfile: etcdCertfile: etcdCAFile: etcdUsername: etcdPassword: help:false version:false autoDetectIPv4:false autoDetectIPv6:false kubeSubnetMgr:true kubeApiUrl: kubeAnnotationPrefix:flannel.alpha.coreos.com kubeConfigFile: iface:[] ifaceRegex:[] ipMasq:true subnetFile:/run/flannel/subnet.env subnetDir: publicIP: publicIPv6: subnetLeaseRenewMargin:60 healthzIP:0.0.0.0 healthzPort:0 charonExecutablePath: charonViciUri: iptablesResyncSeconds:5 iptablesForwardRules:true netConfPath:/etc/kube-flannel/net-conf.json setNodeNetworkUnavailable:true} W1204 11:18:34.968505 1 client_config.go:608] Neither --kubeconfig nor --master was specified. Using the inClusterConfig. This might not work. I1204 11:18:35.156069 1 kube.go:120] Waiting 10m0s for node controller to sync I1204 11:18:35.156105 1 kube.go:378] Starting kube subnet manager I1204 11:18:36.158592 1 kube.go:127] Node controller sync successful I1204 11:18:36.158622 1 main.go:237] Created subnet manager: Kubernetes Subnet Manager - master02.fe.me I1204 11:18:36.158628 1 main.go:240] Installing signal handlers I1204 11:18:36.158733 1 main.go:459] Found network config - Backend type: vxlan I1204 11:18:36.158794 1 main.go:651] Determining IP address of default interface I1204 11:18:36.159132 1 main.go:698] Using interface with name enp0s3 and address 192.168.1.86 I1204 11:18:36.159172 1 main.go:720] Defaulting external address to interface address (192.168.1.86) I1204 11:18:36.159177 1 main.go:733] Defaulting external v6 address to interface address (<nil>) I1204 11:18:36.159246 1 vxlan.go:137] VXLAN config: VNI=1 Port=0 GBP=false Learning=false DirectRouting=false E1204 11:18:36.159477 1 main.go:325] Error registering network: failed to acquire lease: node “master02.fe.me” pod cidr not assigned W1204 11:18:36.159624 1 reflector.go:424] github.com/flannel-io/flannel/subnet/kube/kube.go:379: watch of *v1.Node ended with: an error on the server (“unable to decode an event from the watch stream: context canceled”) has prevented the request from succeeding I1204 11:18:36.159633 1 main.go:439] Stopping shutdownHandler…

I’ve been hit by the same issue on a 1.19.4 cluster. looks like for some reasons, kube-controller-manager doesn’t update the podCIDR of all the nodes (in my 3+3 setup, only 5 out of 6 get a podCIDR), so flannel fails to start on the nodes where it’s missing: failed to acquire lease: node "node-name" pod cidr not assigned

my kubeadm-config configMap looks like this:

    networking:
      dnsDomain: cluster.local
      podSubnet: 10.46.128.0/21
      serviceSubnet: 192.168.1.0/24

and flannel’s one like this:

apiVersion: v1
data:
  cni-conf.json: |
    {
      "name": "cbr0",
      "cniVersion": "0.3.1",
      "plugins": [
        {
          "type": "flannel",
          "delegate": {
            "hairpinMode": true,
            "isDefaultGateway": true
          }
        },
        {
          "type": "portmap",
          "capabilities": {
            "portMappings": true
          }
        }
      ]
    }
  net-conf.json: |
    {
      "Network": "10.46.128.0/21",
      "Backend": {
        "Type": "vxlan",
        "VNI": 1,
        "Port": 8472
      }
    }
kind: ConfigMap

It does work on a 1.18 cluster though (all nodes got podCIDR field), with the same version of flannel, so I’m not sure it’s a flannel issue