Control Plane metrics aren't being scraped correctly when BB deployed https://prometheus.bigbang.dev/targets
When I install BigBang to RKE cluster prometheus isn't able to scrape the following:
monitoring/monitoring-monitoring-kube-kube-controller-manager/0 (0/1 up)
monitoring/monitoring-monitoring-kube-kube-etcd/0 (0/1 up)
monitoring/monitoring-monitoring-kube-kube-proxy/0 (0/4 up)
monitoring/monitoring-monitoring-kube-kube-scheduler/0 (0/1 up)
Using version: v1.18.12+rke2r2
BB UHC: 1.2.0
If you exec into the control plane pods and have them
curl :<prometheus_port>/metrics
you can see they are exporting metrics, it's just prometheus server gets error trying to curl it with connection reset/refused messages.
Activity
changed title from Control Plane isn't scraping correctly https://prometheus.bigbang.dev/targets to Control Plane metrics aren't being scraped correctly when BB deployed https://prometheus.bigbang.dev/targets
This is a known issue, RKE2 has some docs to go in and configure each of those pods to no listen on localhost only but exposes a lot more information on those ports so we have marked it as a known issue for a while and customers are free to find that info themselves and implement.