1. Introduction
  2. Tutorials
  3. 1. Getting Started
  4. 2. Installation on GNU/Linux
  5. 3. Installation on Windows
  6. 4. Docker-compose
  7. 5. Compilation for GNU/Linux
  8. 6. Compilation for Windows
  9. 7. Power consumption of a Kubernetes cluster with scaphandre, prometheus and grafana
  10. How-to guides
  11. 8. Propagate power consumption metrics from hypervisor to virtual machines (Qemu/KVM)
  12. 9. Get process-level power consumption in my grafana dashboard
  13. 10. Install Scaphandre with only Prometheus-push exporter compiled, for Prometheus Push Gateway, on RHEL 8 and 9
  14. Explanations
  15. 11. Explanations about host level power and energy metrics
  16. 12. How scaphandre computes per process power consumption
  17. 13. Internal structure
  18. 14. About containers
  19. 15. About RAPL domains
  20. References
  21. 16. Metrics available
  22. 17. JSON exporter
  23. 18. Prometheus exporter
  24. 19. Qemu exporter
  25. 20. Riemann exporter
  26. 21. Stdout exporter
  27. 22. Warp10 exporter
  28. 23. MSR_RAPL sensor
  29. 24. PowercapRAPL sensor
  30. 25. MSRRAPL sensor
  31. Why this project ?
  32. Compatibility
  33. Troubleshooting
  34. Contributing guide
  35. External references you may be interested in

Scaphandre documentation

Additional references for documentation

  • /proc/stat explained
  • Gathering CPU utilization from /proc/stat
  • proc filesystem documentation
  • CPU usage on Linux
  • Using RAPL to read PP0 and DRAM energy on Haswell
  • RAPL reference
  • How to measure linux performance avoiding most typical mistakes: CPU
  • How to calculate cpu utilization

Powercap/RAPL source code in the kernel

  • arch/x86/events/intel/rapl.c
  • drivers/powercap

Topics to investigate

  • PSI
  • Linux PSI by example
  • PSI - Pressure Stall Information
  • How to measure linux performance avoiding most typical mistakes: memory