12/21/2022, 12:18 PM
Hi. I am not sure is it correct channel for question but if not at least I’ll get correct channel I hope. Working with Rancher 2.7 and rancher-monitoring version: 101.0.0+up19.0.3 is installed. Initially when opening pod from list and opening Pod’s metrics tab graphs are empty. But going under deployment and opening metrics under deployments graphs with data are there. Investigated it by myself and discovered if I’ll remove container!=“” filter from grafana under Rancher/Pod (containers) then in Rancher UI under pod metrics data will be produced. When I try do debug in example prometheus query sum(rate(container_cpu_system_seconds_total{container!=“POD”,namespace=~“$namespace”,pod=~“$pod”,container!=“”}[$__rate_interval])) and remove container!=“” it will work. Prometheus query explorer even does not provide in case of container_cpu_system_seconds_total an option to choose container filter.