Automating Performance Tuning with Machine Learning

Automating Performance Tuning with Machine Learning

USENIX via YouTube Direct link

The optimization goals & constraints

12 of 19

12 of 19

The optimization goals & constraints

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Automating Performance Tuning with Machine Learning

Automatically move to the next video in the Classroom when playback concludes

  1. 1 Intro
  2. 2 SREs care about efficiency and performan
  3. 3 Tuning system configuration matters...
  4. 4 but it is getting harder and harder
  5. 5 Key requirements for a new approach
  6. 6 ML techniques for smart exploration
  7. 7 ML enables automated performance tuning
  8. 8 and a new performance tuning process
  9. 9 The target system: Online Boutique
  10. 10 Use Case: optimizing cost of K8s microservices while ensuring reliability
  11. 11 The reference architecture
  12. 12 The optimization goals & constraints
  13. 13 Best configuration found by ML in 24H improves cost efficiency by 77%
  14. 14 Best config: optimal resources assigned to microservices
  15. 15 Best config: higher performance & efficiency for the overall service Baseline vs Best Service throughout Baseline vs Best Service po response time
  16. 16 Use Case: maximizing service performance & efficiency with JVM tuning
  17. 17 Best config: +28% throughput, and meeting SLOS
  18. 18 Best config: optimal JVM options 8
  19. 19 Key takeaways

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.