Action Runner Controller Impact on Parallel execution of WorkFlow

[ajitkumarnayak1976]

i have K8 controller for Github Actions self-hosted runners. i would like to analysis on estimating memory, CPU and Kubernetes behavior to Ephemeral Runner load while executing mono/multi GitHub Workflow. The analysis will depict the behavior, Usage and Impact of Action Runner Controller while executing multiple workflow in parallel execution per LOB as defined by Ephemeral Runner.  The estimate would provide below details

1. Related to memory, CPU requirement at LOB level (to figure out for scaling)
2. What is action runner controller have options for turning on/off sizing and memory
3. how to scale for CPU/memory using Action Runner Controller
   What would be best option to evaluate it or criteria/ attribute to measure it.