EKS Experiments #3: Chaos Testing with AWS FIS (What Broke First?)

EKS can look “healthy” right up until you inject failure.

In this EKS Experiments episode, I run a series of AWS Fault Injection Service (AWS FIS) chaos tests against a live EKS cluster under load to see which experiments pass and which ones push our cluster to its limits. You’ll get real metrics, real outcomes, and what I’d change to make the system more resilient.

✅ Github Repo: https://github.com/robertdippolito/AWS-Fault-Injection-Simulator-Experiment-Templates

What you’ll learn

What AWS FIS is (and why it’s useful for EKS chaos engineering)

My test setup: EKS + load + the experiments I ran

PASS vs FAIL results (what broke first + why)

Real impact on the system (health + latency)

Practical takeaways to harden your cluster

Chapters

00:28 AWS FIS + chaos engineering on EKS
00:48 Test setup (cluster + load)
01:40 Experiments CPU
02:49 Experiments Memory
04:04 Experiments Network Latency
05:22 Experiments Pod Deletion
06:21 Experiment Wrap up

💬 Question: Which failure would you test first on EKS — CPU, memory, network, pod deletion?

#EKS #Kubernetes #AWSFIS #ChaosEngineering #AWS #DevOps #SRE #PlatformEngineering

Видео EKS Experiments #3: Chaos Testing with AWS FIS (What Broke First?) канала RobOps