1. Proposed title of this feature request

Currently there are PSI metrics at node level [1], but we'd like metrics at container level. There is an open PR in cAdvisor "Support for exposing PSI metrics" [2] , which will be merged in some days, as runc dependcy has already been merged [3] . And there is a KEP "Support PSI based on cgroupv2" [4] .

[1]https://access.redhat.com/solutions/6987181

   |- node_pressure_cpu_waiting_seconds_total
   |- node_pressure_io_stalled_seconds_total
   |- node_pressure_io_waiting_seconds_total
   |- node_pressure_memory_stalled_seconds_total
   |- node_pressure_memory_waiting_seconds_total
[2]https://github.com/google/cadvisor/pull/3083

[3]https://github.com/kubernetes/kubernetes/pull/128276

[4]https://github.com/kubernetes/enhancements/issues/4205

2. What is the nature and description of the request?

With this RFE we'd like to ensure that Openshift will also expose the metrics, and have a follow-up of the feature to know the version where the feature will be available.

3. Why does the customer need this? (List the business requirements here)

Missing PSI metrics at container level

4. List any affected packages or components.

Monitoring