Chaos Mesh

Chaos Mesh is a cloud-native chaos engineering platform that helps organizations identify and fix system vulnerabilities by injecting controlled failures into Kubernetes environments. Originally developed at PingCAP (the company behind TiDB) in 2019 and now a CNCF Incubating project, Chaos Mesh has over 7,000 stars as of 2026. Chaos Mesh runs as a Kubernetes-native tool, using custom resource definitions (CRDs) to define chaos experiments as declarative Kubernetes objects, making chaos engineering as simple as applying a YAML file. Key features include: comprehensive fault injection types including pod-level chaos (pod-kill, container-kill, pod-failure for simulating node failures), network chaos (network delay, loss, duplication, corruption, partition, bandwidth restriction for testing network resilience), I/O chaos (delaying or erroring file system operations like read, write, open, and fsync), DNS chaos (returning random or error responses for DNS queries), stress testing (CPU and memory stress via stress-ng), time skew (shifting the system clock for testing time-sensitive logic), HTTP chaos (delaying or modifying HTTP responses), JVM chaos (injecting GC pressure, method exceptions, and return values in Java applications via Byteman), kernel chaos (injecting kernel faults), and disk chaos (filling disk space or read/write errors), workflow orchestration (defining sequential, parallel, and conditional chaos experiments via Workflow CRD), experiment scheduling (running chaos experiments at specific times or intervals), automated workflows for continuous chaos testing, observability via Chaos Dashboard (a web UI for managing and monitoring experiments), and integration with CI/CD pipelines.