Rancher Ingress isn't bound on port 80/443

I’m running on Centos 7.4, with RKE 1.3.4 and picked the hyperkube 1.21.7 as my system_image in the cluster.yml

RKE comes up fine (3 control plane / etcd and 3 workers) on a total of 6 nodes. rke up completed without errors. I have deployed the cert-manager 1.7.1 , also without errors.

When I deployed rancher 2.6.3 using helm all the pods come up, but I can’t connect to the UI. It seems that port 80, and 443 on the worker nodes where the ingress is deployed are not bound, and I can’t understand why. I’ve used pretty much basic options in the cluster.yml and not overwritten any ports or anything like that. Ingress is installed, you can see the controller pods, and the ingress resource lists the correct domain that it is listening to and ports 80, 443

I didn’t get any kind of errors like 80,443 are already in use, or permissions or anything like that. Any ideas why ingress is ‘up’ but not up?


Just to clarify:
The Ingress does not bind anything, the Ingress Controller does. Ingresses are more or less just a set of rules.

Where is your Ingress Controller and how does it look?

Hey Akito,

The ingress controller is deployed on the worker nodes and there is 1 pod on each node, they are all running. But I’m expecting to see ports 80 and 443 with some kind of Nginx process running but I don’t