Hey folks, new to the channel. I'm running k3d v5...
# general
s
Hey folks, new to the channel. I'm running k3d v5.8.3 k3s version v1.31.5-k3s1. I'm trying to get rancher ui monitoring dashboards working. The ui shows unavailable. I set CATTLE_PROMETHEUS_METRICS to true for the rancher deployment. I'm guessing it's a port I need to open. I'm seeing 404s in devtools with URLS like: https://localhost:30081/api/v1/namespaces/cattle-monitoring-system/services/http:ran[…]g-grafana:80/proxy/api/dashboards/uid/rancher-cluster-nodes-1 Is the the right channel or should I use another channel? Thanks for any hints
b
Rancher typically forwards/proxies the ports right. It looks like it's expecting port
30081
to be open and it's not. It's possible that there's something weird going on with the proxy forward to the port 80 service, but typically we never had to touch that.
That being said, we're not running the rancher ui through a non standard port like you are, we're using port 443. Let me double check the HAProxy config
Yep, ports 80 443 6443 and 9345
s
thanks, I'll give it a whirl
b
I should mention that this is with k3s and not k3d.
s
okay, yeah k3d adds a twist for sure.
I think my issue is that I installed monitoring via kube-prometheus-stack and installed it into monitoring namespace vs cattle-monitoring-system