Kubernetes Performance Tuning Workshop

on YouTube ↗02h 26m

" summary

Hands‑on workshop shows a three‑node k3d k3s cluster with metric server preinstalled and runs live demos: a Flask test app with 50m CPU causing throttling makes 1s work turn into ~25s under 20 concurrent requests, removing CPU limits restores latency; an ephemeral volume capped at 100KB causes pod eviction during dd writes; k6 operator runs distributed load tests; Vertical Pod Autoscaler and Perfect Scale recommendations generate resource suggestions and can automate resizing, with caveats about volatility and admission conflicts.

AI-written summary · about

" chapters (26)

1. 0:00🚀 Introduction and Overview of the Kubernetes Performance Tuning Workshop
2. 3:00Signing into the Virtual Labs for Hands-On Participation
3. 6:00Introduction to Kubernetes performance tuning and the concept of optimization
4. 10:00Understanding the importance of resource allocation and utilization in Kubernetes
5. 15:00Overview of Kubernetes architecture: control plane vs. data plane
6. 20:00Introduction to the virtual environment (stgo) and accessing Kubernetes clusters
7. 24:00Setting up Perfect Scale and explaining its benefits for Kubernetes optimization
8. 28:00Overview of Kubernetes performance metrics and how to gather data
9. 33:00Hands-on exercise: Installing Perfect Scale and checking Kubernetes resource metrics
10. 40:00Optimizing container images for better performance
11. 50:00CPU and memory optimization: measuring and adjusting resource allocation
12. 55:00Exploring Kubernetes storage optimization strategies
13. 1:00:00Introduction to Kubernetes' metric server and how it aids in resource optimization
14. 1:05:00Measuring pod resource usage and understanding metrics
15. 1:10:00Overview of the metric server and its role in scaling applications in Kubernetes
16. 1:15:00How to manage resource allocation through Kubernetes requests and limits
17. 1:20:00Setting CPU and memory limits, and understanding pod quality of service
18. 1:35:00Understanding how CPU limits impact performance and avoiding throttling
19. 1:45:00Managing ephemeral storage and handling disk usage in Kubernetes
20. 1:50:00Scaling Kubernetes workloads effectively using vertical and horizontal scaling
21. 2:00:00Advanced scaling techniques and their impact on Kubernetes performance
22. 2:05:00Automation of resource allocation using Perfect Scale and Kubernetes Vertical Pod Autoscaler (VPA)
23. 2:10:00How to optimize workloads with Perfect Scale and integrate it with your Kubernetes cluster
24. 2:15:00Running load tests on Kubernetes applications with k6 and analyzing results
25. 2:20:00Reviewing best practices for continuous performance tuning and resource optimization
26. 2:25:00Final thoughts on Kubernetes performance tuning and using Perfect Scale to automate the process

" related talks

• 34mPanel: Improving and Managing Kubernetes at ScaleXiang Li, Corin Dwyer · KubeCon + CloudNativeCon North America 2019
• 47mOptimizing Kubernetes Resource Requests/Limits for Cost-Efficiency and LatencyHenning Jacobs · HighLoad++ Moscow 2018
• 01h 22mWorkshop: Optimizing Kubernetes Cluster Scaling for Adv...Shivay Lamba, Shivanshu Raj Shrivastava · AI_dev: Open Source GenAI & ML Summit Europe 2024
• 31mLoad Testing Kubernetes: How to Optimize Your Cluster Resource Allocation in Production [I]Harrison Harnisch · CloudNativeCon + KubeCon Europe 2017
• 32mOptimizing Application Performance on KubernetesDinakar. Guniguntala · Kubernetes Community Days Chennai 2022
• 25mTune It Up! Enabling Low Latency in Kubernetes ClustersFrancesco Romani, Yanir Quinn · KubeCon + CloudNativeCon North America 2020 Virtual
• 44mFinOps at scale: Lessons from scaling Kubernetes workloadsRicardo Magalhães · JNation 2026
• 36mUnderstanding Scalability and Performance in the Kubernetes MasterXingyu Chen, Fansong Zeng · KubeCon + CloudNativeCon + Open Source Summit China 2019