Review and redesign the original rteval-runner topology management logic and the per-CPU/percentage-based allocation script (currently implemented separately in Bash). Separate these concerns into well-defined layers, ensuring the following:

• Dedicated Monitoring CPUs:
  Design the system to allow reserving the first X CPUs specifically for monitoring tools (distinct from housekeeping CPUs).

• Partition Scalability:
  Ensure the scripts and overall architecture support seamless movement between isolated and load partitions, enabling advanced, profile-driven simulations.

• Profile "Knob" Mechanism:
  Implement a mechanism to easily adjust simulation profiles using a "knob effect" (e.g., percentage or CPU count), supporting at least the three main scenarios discussed.

Stretch Goal:
If time permits, begin prototyping or implementing core elements of the new design.