Skill

chaos-engineering

Chaos Engineering expertise from Netflix - the company that invented Chaos Monkey. Build resilient systems through controlled failure injection. Use when implementing resilience testing, chaos experiments, or building fault-tolerant systems. Learn from the pioneers of chaos engineering.

npx claudepluginhub duylinhdang1998/claude-template-agent --plugin vfm-agent-company

Tool Access

This skill uses the workspace's default tool permissions.

Preview

**Expert**: Marcus Johnson (Netflix, 14 years)

Supporting Assets

references/chaos-monkey.mdreferences/experiments.md

SKILL.md

Similar Skills

skill-lookup

159.9k

Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.

prompts.chat

prompt-lookup

159.9k

Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.

prompts.chat

next-compile

139.2k

Checks Next.js compilation errors using a running Turbopack dev server after code edits. Fixes actionable issues before reporting complete. Replaces `next build`.

1 file

vercel-next-js-2

Stats

Parent Repo Stars0

Parent Repo Forks0

Last CommitApr 16, 2026

Actions

View Source View Plugin View on GitHub View README

Chaos Engineering - Netflix Chaos Monkey Expertise

Expert: Marcus Johnson (Netflix, 14 years) Level: 10/10 - Netflix invented Chaos Engineering

Overview

Chaos Engineering from Netflix - created Chaos Monkey and pioneered failure testing in production. Learn how Netflix achieves 99.999% uptime by intentionally breaking systems.

Why Netflix Created Chaos Engineering (2010)

The Problem:

Migrating from monolith to cloud microservices
Hundreds of services with complex dependencies
Traditional testing didn't catch production failures
"Works in dev" ≠ "Works at scale in production"

The Solution: Chaos Monkey (2011)

Randomly terminate EC2 instances in production
Force engineers to build resilient services
Discover weaknesses before customers do
Cultural shift: Assume everything fails

Core Principles

Define steady state - Normal system behavior
Hypothesize - Predict system response to failures
Inject chaos - Controlled blast radius
Measure and learn - Track impact, improve resilience

Chaos Experiments Workflow

Measure baseline - Capture normal metrics
Define hypothesis - "IF failure X, THEN system remains stable BECAUSE Y"
Set abort conditions - Safety thresholds
Inject failure - Start small (1% instances)
Observe impact - Monitor metrics
Analyze results - Did hypothesis hold?
Improve resilience - Fix weaknesses found

Netflix Simian Army Tools

Chaos Monkey - Terminate random instances in production
Latency Monkey - Inject artificial delays into services
Chaos Kong - Simulate entire AWS region failure

Building Resilience (Netflix Patterns)

Redundancy - Multiple instances, multi-AZ, multi-region
Graceful degradation - Fallbacks for failures
Circuit breakers - Stop cascading failures
Auto-healing - Kubernetes restarts failed pods

Netflix's Chaos Maturity Model

Level 1: Ad-hoc manual testing Level 2: Scheduled chaos experiments (GameDays) Level 3: Automated chaos in pre-prod Level 4: Continuous chaos in production ← Netflix is here Level 5: Chaos as a service for all teams

Best Practices

Start small - 1% instances, single service, dev/staging first
Define success criteria - Measurable steady state metrics
Automate everything - Execution, metrics, rollback
Learn and improve - Document findings, fix weaknesses

Related Skills

microservices - Resilient architecture
sre-practices - Reliability engineering
kubernetes-expert - Container resilience
observability - Measure chaos impact

Last Updated: 2026-02-03 Expert: Marcus Johnson (Netflix, 14 years) - 99.999% uptime at 260M+ subscribers