I'm setting up Rancher with K3s and I'm unsure whether Rancher is still actively maintained and commonly used. During setup I'm seeing connection-refused errors mentioning port 80, even though I expected the K3s API server to use port 6443. I also changed a 172.0.0.1 address to the machine's IP, but that didn't resolve the problem. Rancher eventually generated a URL resembling rancher.io.com, but I couldn't open that either. The ingress currently shows Rancher using the Traefik class, with hosts such as rancher.my.org, addresses on my local network, and ports 80 and 443. What should I check, and how should the Rancher hostname and ports be configured?
3 Answers
The ingress output suggests Rancher is being exposed through Traefik on ports 80 and 443, and the hostname is `rancher.my.org`. You need to make that hostname resolve to the node or load-balancer address shown in the `ADDRESS` column. For a local setup, you can temporarily add the hostname and IP to your computer’s hosts file, or configure local DNS. Then browse to `https://rancher.my.org`, accepting that a self-signed certificate may produce a warning. Don’t replace an internal address with a random IP unless you know which component generated it.
Yes, Rancher is still actively maintained and is commonly used with K3s. The ports are easy to mix up: 6443 is the Kubernetes API server, while Rancher’s web interface normally uses HTTP on 80 and HTTPS on 443 through the ingress controller. A refused connection usually means the ingress, load balancer, firewall, DNS, or another service listening on those ports needs investigation. Check the Rancher and ingress-controller pods, then inspect the node with `ss -tulpn` to see what is actually listening on 80 and 443.
Would the official support site be the right place to report this if I still can’t determine which component is refusing the connection?
For a Helm-based installation, the hostname is generally configured through Rancher’s ingress settings rather than by changing the K3s API port. Use the same fully qualified hostname consistently during installation, make sure it resolves correctly, and confirm that the Rancher pods and ingress resources are healthy with commands such as `kubectl get pods -n cattle-system` and `kubectl describe ingress rancher -n cattle-system`. Also check whether another web server is already occupying ports 80 or 443 on the node.
The ingress shows two local-network addresses for the hostname, with ports 80 and 443. I’ll verify that the hostname resolves to one of those addresses and that the browser is using HTTPS.