CHAOS-ENGINEERING

SRE / DevOps concept Chaos engineering.

Definition

Chaos engineering is the deliberate and controlled introduction of failures in production (Pod kill, injected latency, network loss, dependency down) to verify the hypothesis that the system holds up. Five principles formalised by Netflix in 'Principles of Chaos Engineering' (2017). Tools: Chaos Monkey, Chaos Mesh, Litmus, Gremlin.

Origin

Practice born at Netflix in 2010 (Chaos Monkey), publicly formalised in 2017.

Example in context

EDI game day: kill the XSLT mapping pod during peak New Year retail INVOIC flow, verify Kafka consumer failover.

Related terms

• Game day — planned chaos exercise.