I have created nginx deployment and nginx service(ClusterIP) to access nginx pod. But not able to access pod through cluster IP across nodes other than node where pod is scheduled.
I tried looking for IPtable too. But do not DNAT entry over there.
root@kdm-master-1:~# k get all -A -o wide |grep nginx
default pod/nginx-6db489d4b7-pfkm9 1/1 Running 0 3h16m 10.244.1.3 kdm-worker-1 <none> <none>
default service/nginx ClusterIP 10.102.239.131 <none> 80/TCP 3h20m run=nginx
default deployment.apps/nginx 1/1 1 1 3h32m nginx nginx run=nginx
default replicaset.apps/nginx-6db489d4b7 1 1 1 3h32m nginx nginx pod-template-hash=6db489d4b7,run=nginx
IP table:
root@kdm-master-1:~# iptables -L -t nat|grep nginx
KUBE-MARK-MASQ tcp -- !10.244.0.0/16 10.102.239.131 /* default/nginx:80-80 cluster IP */ tcp dpt:http
KUBE-SVC-OVTWZ4GROBJZO4C5 tcp -- anywhere 10.102.239.131 /* default/nginx:80-80 cluster IP */ tcp dpt:http
# Warning: iptables-legacy tables present, use iptables-legacy to see them
Please advice how can I resolve it?
set net.ipv4.ip_forward=1 in /etc/sysctl.conf
run sysctl --system
This will resolve the issue and one will be able able to access the pod from any node.