Flipkart won the CNCF End User Case Study Contest at KubeCon + CloudNativeCon India 2026 for building a centralized, multi-tenant chaos engineering platform on Kubernetes using LitmusChaos. Their central reliability engineering team engineered four custom extensions including a hybrid multi-tenant architecture, a DaemonSet-based high-availability injection model, a Script Runner fault for dynamic target selection, and support for legacy VM workloads. About 90% of chaos experiments run in staging environments ahead of high-traffic festive sales events. The team also contributed five upstream fixes back to the LitmusChaos project, addressing database index issues, duplicate-name validation, and workflow configuration bugs. Future plans include integrating chaos testing into CI/CD pipelines and open-sourcing their DaemonSet HA injection model.

5m read timeFrom cncf.io
Post cover image
2 Impressions