Exposing metrics in Incus: Is it possible to get metrics (incus_cpu_seconds_total) of all the CPU (say 4) of a container?

Incus
1

My observation:
When i stress a container incus exec cnt0 – stress-ng --cpu 4 --timeout 120. and check CPU Affinity inside the container, using the command i get the below output, incus exec cnt0 – bash -c ‘for pid in $(pgrep stress-ng); do taskset -cp $pid; done’
pid 1988’s current affinity list: 0-3
pid 1989’s current affinity list: 0-3
pid 1990’s current affinity list: 0-3
pid 1991’s current affinity list: 0-3

It means stress is distributed among the CPU but when i expose metrics, only CPU 0 metrics is visible. Please find below the output incus_cpu_seconds_total{cpu=“0”,mode=“system”,name=“cnt0”,project=“default”,type=“container”} 1202.286404
incus_cpu_seconds_total{cpu=“0”,mode=“user”,name=“cnt0”,project=“default”,type=“container”} 1217.502465

it just shows cpu=“0” details ! What can i do to see all the CPU metrics, when i get the metrics exposed by incus ?

2

In Grafana I’m using this query to get CPU usage relative to the number of CPUs available in a container:

sum by (name) (rate(incus_cpu_seconds_total{mode=~"user|system", instance="${incus_instance}"}[$__rate_interval])) / count without (cpu) (sum by (name, cpu) (incus_cpu_seconds_total{mode=~"user|system", instance="${incus_instance}"} > 10))

It might be helpful to you.