Chaos Monkey
By Netflix
Chaos Monkey is a chaos engineering tool, originally built by Netflix, that randomly terminates production instances to test whether a system can tolerate unexpected failures without degrading service.
Definition
Chaos Monkey is a chaos engineering tool, originally built by Netflix, that randomly terminates production instances to test whether a system can tolerate unexpected failures without degrading service.
Overview
Chaos Monkey was created by Netflix's engineering team in the early 2010s as Netflix migrated to AWS, based on the premise that instance failures in the cloud are inevitable and that the best way to build confidence a system can survive them is to trigger those failures deliberately and often, during business hours, rather than waiting to discover weaknesses during a real, uncontrolled outage. It runs continuously, randomly selecting virtual machine instances within a defined service group and terminating them, forcing engineering teams to build services that tolerate individual instance loss gracefully — through redundancy, health checks, and automatic failover — rather than as an afterthought. Chaos Monkey was the first and simplest member of Netflix's broader Simian Army, a suite of chaos tools that also included Latency Monkey for injecting artificial delays and Conformity Monkey for flagging instances that didn't follow best practices; most of the Simian Army has since been retired or folded into Netflix's later Chaos Automation Platform. Netflix open-sourced Chaos Monkey in 2012, and it became one of the founding artifacts of the broader chaos engineering discipline formalized by the Principles of Chaos Engineering. While Chaos Monkey itself narrowly targets instance termination, it inspired a much broader ecosystem of chaos engineering tools — including Gremlin, LitmusChaos, and the Chaos Toolkit — that extend the same philosophy to network partitions, resource exhaustion, and dependency failures. Practicing Chaos Monkey-style testing is closely tied to the discipline of Site Reliability Engineering, which treats proactively surfacing failure modes as essential to operating reliable systems at scale.
Key Features
- Randomly terminates production instances to test failure tolerance
- Runs during business hours so engineers are present to observe and respond
- Originated as the first tool in Netflix's Simian Army chaos suite
- Open source, encouraging adoption of chaos engineering practices industry-wide
- Forces services to be designed for redundancy and automatic failover
- Configurable scope, limiting terminations to specific service groups
- Foundational influence on later chaos engineering tools and platforms