Watch and track your favorite playlist.
Curated by: CNCF [Cloud Native Computing Foundation] (43 videos)
Achieving Fault Tolerance in Istio with Observability-Driven Load Management - Tanveer Gill, FluxNinja Service mesh technologies like Istio have revolutionized microservices communication. Yet, ensuring consistent reliability remains a challenge. Metastable failures like cascading failures and retry storms, highlight the limitations of traditional mitigation strategies such as circuit breakers and rate-limiting, especially when adapting to dynamic service conditions. In this context, observability-driven automation is required. Aperture is an open-source load management system, that introduces adaptive service protection, workload prioritization, and more. Aperture Agent operates as a Sidecar or as a DaemonSet and integrates with Istio using Envoy's External Authorization API to gain traffic insights to make informed decisions that safeguard against failures. This session will showcase Aperture in a real-world deployment where it is used to protect multi-tenant databases such as Apache Druid and PostgreSQL from overloads by adaptively scheduling GRPC and GraphQL traffic.