kepler container_package does not equal total of kepler_process_package

[sthaha]

Steps to reproduce on a Baremetal

• deploy kepler
• run a container that stresses the cpu
• curl /metrics
• grep for container_package_joules where container_id is not system_processes or kernel_processes
• grep for kepler_process_package_joules related to that container id
• find sum of rate the values of processes
• see if the total is different to container_package

Expected: there shouldn't be any significant difference
Actual: The difference quite large and grows over time

Using Prometheus

sum by(job, container_id)(rate(kepler_container_package_joules_total{job=~"host|dev|latest", container_id!~"system_processes|kernel_processes"}[$rate])) 

• red is the containers' power
• yellow is the processes power summed up.

This also means the pod's consumption reported is also wrong