kubernetes: IPVS error log occupation with externalTrafficPolicy: Local option in Service

What happened:

IPVS error log occupation with externalTrafficPolicy: Local

$ dmesg
...
[23709.680898] IPVS: rr: TCP 192.168.0.52:80 - no destination available
[23710.709824] IPVS: rr: TCP 192.168.0.52:80 - no destination available
[23832.428700] IPVS: rr: TCP 127.0.0.1:30209 - no destination available
[23833.461818] IPVS: rr: TCP 127.0.0.1:30209 - no destination available
...

What you expected to happen:

Do not print the error log.

How to reproduce it (as minimally and precisely as possible):

$  kubectl get nodes -o wide
NAME     STATUS   ROLES                  AGE   VERSION   INTERNAL-IP    EXTERNAL-IP   OS-IMAGE             KERNEL-VERSION     CONTAINER-RUNTIME
node01   Ready    control-plane,master   13d   v1.20.5   192.168.0.61   <none>        Ubuntu 18.04.4 LTS   5.4.0-64-generic   docker://19.3.12
node02   Ready    <none>                 13d   v1.20.5   192.168.0.62   <none>        Ubuntu 18.04.4 LTS   5.4.0-64-generic   docker://19.3.12

$ kubectl get service
NAME              TYPE           CLUSTER-IP       EXTERNAL-IP      PORT(S)        AGE
kubernetes        ClusterIP      10.96.0.1        <none>           443/TCP        13d
nginx-ipv4        LoadBalancer   10.96.82.45      192.168.0.52     80:30209/TCP   54m

$ kubectl get service nginx-ipv4 -o yaml
...
spec:
  clusterIP: 10.96.82.45
  clusterIPs:
  - 10.96.82.45
  externalTrafficPolicy: Local
  healthCheckNodePort: 31583
  ipFamilies:
  - IPv4
  ipFamilyPolicy: SingleStack
  ports:
  - nodePort: 30209
    port: 80
    protocol: TCP
    targetPort: 80
  selector:
    run: nginx
  sessionAffinity: None
  type: LoadBalancer
status:
  loadBalancer:
    ingress:
    - ip: 192.168.0.52

$ kubectl get pod -o wide
NAME                                  READY   STATUS    RESTARTS   AGE   IP                NODE     NOMINATED NODE   READINESS GATES
nginx-658f4cf99f-654bc                1/1     Running   0          55m   192.167.140.91    node02   <none>           <none>
nginx-658f4cf99f-jp27p                1/1     Running   0          55m   192.167.140.92    node02   <none>           <none>
nginx-658f4cf99f-ld9th                1/1     Running   0          55m   192.167.140.82    node02   <none>           <none>

(node01) $ ipvsadm -L
...
TCP  192.168.0.52:80 rr
TCP  192.168.0.61:30209 rr
...

(node01) $ curl 192.168.0.52:80
(node01) $ curl localhost:30209

[23709.680898] IPVS: rr: TCP 192.168.0.52:80 - no destination available
[23710.709824] IPVS: rr: TCP 192.168.0.52:80 - no destination available
[23832.428700] IPVS: rr: TCP 127.0.0.1:30209 - no destination available
[23833.461818] IPVS: rr: TCP 127.0.0.1:30209 - no destination available

As you can see, “nginx-ipv4” service endpoint are nginx pods in “node02”. Because “nginx-ipv4” service has “externalTrafficPolicy: Local” and node01 doesn’t have nginx pod, node01 doesn’t have IPVS destination for 192.168.0.52:80 and 192.168.0.61:30209. And this makes IPVS occupy the kernel log.

I know this is intended. But I think kube-proxy must set the filter ruless to iptables for 192.168.0.52:80 and 192.168.0.61:30209 to prevent IPVS error log occupation. Is there any plan for this issue??

Anything else we need to know?:

Environment:

• Kubernetes version (use kubectl version):
  • v1.20.5
• Cloud provider or hardware configuration:
  • my local cluster
• OS (e.g: cat /etc/os-release):

NAME="Ubuntu"
VERSION="18.04.4 LTS (Bionic Beaver)"
ID=ubuntu
ID_LIKE=debian
PRETTY_NAME="Ubuntu 18.04.4 LTS"
VERSION_ID="18.04"
HOME_URL="https://www.ubuntu.com/"
SUPPORT_URL="https://help.ubuntu.com/"
BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/"
PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy"
VERSION_CODENAME=bionic
UBUNTU_CODENAME=bionic

• Kernel (e.g. uname -a):
  • Linux node01 5.4.0-64-generic #72~18.04.1-Ubuntu SMP Fri Jan 15 14:06:34 UTC 2021 x86_64 x86_64 x86_64 GNU/Linu
• Install tools:
  • kubeadm
• Network plugin and version (if this is a network-related bug):
• Others: