Docs feedback: recommended Kubernetes setup (node exporter)

[Unb0rn]

URL

https://grafana.com/docs/alloy/latest/reference/components/prometheus/prometheus.exporter.unix/

Component(s)

prometheus.exporter.unix

Feedback

We have a nice doc on k8s-monitoring-helm chart for Grafana cloud.
It would be great to adapt it to onprem installation and modernize the config.

It is pretty strange that the chart mentioned above relies on node-exporter, while Alloy has prometheus.exporter.unix

We also have another useful discussion with solution for cAdvisor and kubelet metric scraping.

It's hard to choose - deploying Alloy as DaemonSet and enabling clustering for everything (ServiceMonitors, kubelet, cAdvisor) except for prometheus.exporter.unix job - is this a good way to go?
Or maybe redeploying Alloy with host namespace and directly scraping localhost ports for kubelet metrics with relabellings is preferred?
Or just stick with standalone node exporter?

I think compatibility with existing dashboards should be a priority

Tip

_{React with 👍 if this issue is important to you.}