Missing metrics on nodes (CPU, RAM) HPA failing

Rancher 2.x
1

Hello,

we are running a Rancher v2.6.11 and seeing issues with metrics. There are no values for the nodes regarding CPU ans RAM. This also causes that HPA is not working as it can not get CPU info. When checking the Targets it shows something like /110%.

In RancherUI it looks like this:

Bildschirm­foto 2023-07-14 um 14.02.34
Bildschirm­foto 2023-07-14 um 14.02.342792×1618 302 KB

The Nodes are all running as KVM virtualized Ubuntu machines in Proxmox. Maybe because of that there are problems getting information? I hope someone can help me out or giving me a hint where I need to check for the issue.

Thanks a lot in advance!

2

It turned out that the metrics-server was just not installed when the cluster was bootstrapped, because it has been disabled manually at installation. I downloaded the configuration for it and applied it manually with kubectl.