Hi,
I installed mettalb in my k8s cluster.
The services got loadbalancer ip address from metallb ip range.
The arp list content the LB IP and mac-address. but the IP address not pinging, not available.
testgitlab-nginx-ingress-controller LoadBalancer 10.101.250.237 10.248.89.110 80:30837/TCP,443:30150/TCP,22:32472/TCP 19m
arp |grep 110
10.248.89.110 ether 00:21:f6:8d:bf:6c C eth0
arping 10.248.89.110 -I eth0 -b
ARPING 10.248.89.110 from 10.248.89.81 eth0
Unicast reply from 10.248.89.110 [00:21:F6:8D:BF:6C] 0.767ms
Unicast reply from 10.248.89.110 [00:21:F6:8D:BF:6C] 0.949ms
ping 10.248.89.110
PING 10.248.89.110 (10.248.89.110) 56(84) bytes of data.
From ... icmp_seq=2 Redirect Host(New nexthop: 10.248.89.110)
--- 10.248.89.110 ping statistics ---
1 packets transmitted, 0 received, 100% packet loss, time 2074ms
kubectl describe configmap -n kube-system kube-proxy | grep ARP
strictARP: false
I tried set strictARP: true parameter, and jube proxy restart, but no change was made.
and a lot of error messages as well
kubectl logs speaker-kqgjm -n metallb-system
{"caller":"speakerlist.go:257","error":"1 error occurred:\n\t* Failed to resolve : lookup : no such host\n\n","expected":1,"joined":0,"msg":"partial join","op":"memberDiscovery","ts":"2021-01-21T17:27:08.000765019Z"}
{"caller":"memberlist.go:236","component":"Memberlist","msg":"[WARN] memberlist: Failed to resolve : lookup : no such host","ts":"2021-01-21T17:28:08.000660645Z"}
{"caller":"speakerlist.go:257","error":"1 error occurred:\n\t* Failed to resolve : lookup : no such host\n\n","expected":1,"joined":0,"msg":"partial join","op":"memberDiscovery","ts":"2021-01-21T17:28:08.000723348Z"}
Idea?
Best Regards
Steven