Tutorial: Debug Your Kubernetes Apps

on YouTube ↗01h 32m

" summary

Cluster debugging focuses on exact failure modes and fixes: IP exhaustion on AWS CNI explains why 240 replicas stop at ~224; deploy CNI metrics helper to monitor pod IPs and alert, then resize subnets or instances. Fix DNS by scaling or tuning CoreDNS (memory, autoPath, node-local cache). Add metric-server for kubectl top, install cluster-autoscaler linked to ASG, use Prometheus+Grafana or CloudWatch Container Insights for alerts, and prefer CSI volumes (EBS/EFS) with StatefulSets to keep data persistent.

AI-written summary · about

" chapters (71)

1. 0:00Introduction
2. 1:00Understanding Kubernetes
3. 1:25Agenda
4. 2:00Cluster Design
5. 2:17Core Kubernetes Components
6. 3:36How to Setup a Kubernetes Cluster
7. 5:27Design Considerations
8. 6:38Amazon Kubernetes Service
9. 7:31QCuddle
10. 7:56Master Node Components
11. 9:20HDD Design
12. 9:43Component Site
13. 10:44Creating a Cluster
14. 11:49Configuration File
15. 13:50What is CI
16. 14:43How CI works
17. 16:16Create a deployment
18. 17:37What went wrong
19. 18:05Troubleshoot
20. 18:51IPAMD
21. 20:00Ichi
22. 20:20Hello World
23. 20:34IP Addresses
24. 22:23CNI Metrics Helper
25. 22:42Create a Policy
26. 22:52Attach to EC2 Instance
27. 23:09Apply YAMO
28. 23:31CloudWatch
29. 23:47Core DNS
30. 24:39Check your DNS pods
31. 25:13Add logging to DNS
32. 25:51Check pod output
33. 26:11Local DNS
34. 26:41Memory Requirements
35. 27:39AutoPath
36. 28:34Cube Kernel
37. 30:03Cube Cuddle
38. 31:30API Server
39. 33:38Common Errors
40. 35:59Create New Cube Config
41. 36:43Cluster Access
42. 38:15I Am Role
43. 39:33CubeCaudill
44. 39:54Get Nodes
45. 40:10Enable Public Access
46. 40:59Pods
47. 41:14Pod Lifecycle
48. 42:35Scale Existing Deployment
49. 43:17Reasons for Pending
50. 44:14Describe a Pod
51. 45:02Get Events
52. 45:47Cubecurl
53. 46:01Memory CPU Requirements
54. 47:29Memory CPU Utilization
55. 48:06Metric Server
56. 48:27Get API Service
57. 48:37CPU Usage
58. 49:26Capacity Memory
59. 50:31Node Capacity
60. 50:55Node Allocatable
61. 51:49Eject Hard and Cube Reserved
62. 52:16Cluster Autoscaler
63. 53:05Autoscaling Group
64. 53:33Create Cluster Autoscaling
65. 53:43Cluster Autoscaling logs
66. 54:08Update Autoscaling group limits
67. 55:07Use case
68. 55:31Solution
69. 56:10Load Balancer
70. 58:11Load Balancer Failure
71. 59:22Load Balancer Logging

" related talks

• 50mTroubleshooting & Debugging Microservices in KubernetesRay Tsang, Robert Kubis · GOTO Chicago 2019
• 35mKubernetes, Lessons LearntKobi Biton · SREday London 2022
• 34mDebugging at Scale in Production - Deep into Your Containers with Kubectl DebugShai Almog · ContainerCon @ Open Source Summit North America 2022
• 50mTroubleshooting & Debugging Microservices in KubernetesRay Tsang, Robert Kubis · GOTO Berlin 2018
• 29mWhen Kubernetes Misbehaves – Pitfalls and Lessons LearnedItiel Shwartz · ContainerDays 2022
• 34mBecoming an Expert of Diagnosing a Kubernetes Control Plane ProblemJoe Betz, Wenjia Zhang · KubeCon + CloudNativeCon China 2018
• 36mKubernetes Confessions: Real Mistakes We’ve All Made (and What They Teach Us)Shashank Pai · DevConf.IN 2026
• 35mKlustered: Kubernetes Debugging LiveDavid Flanagan · SREday London 2024