Balancing Cost and Quality in OpenTelemetry: An Evaluation of Sampling Policies

Explore the critical balance between cost and quality in OpenTelemetry through an in-depth evaluation of sampling policies in this informative conference talk. Delve into the challenges of distributed tracing in microservice architectures, where massive volumes of trace data can lead to significant storage and cost implications. Discover how OpenTelemetry's various sampling strategies can help mitigate resource usage while maintaining data quality. Gain insights into quantitative analyses of different sampling policy combinations, assessing their impact on network I/O, memory, and storage requirements. Learn about innovative sampling approaches designed to capture edge cases and understand how to tailor sampling systems to your specific business needs. Walk away with practical knowledge on the costs and benefits of different sampling policies, empowering you to make informed decisions for optimizing your observability infrastructure.

Balancing Cost and Quality in OpenTelemetry: An Evaluation of Sampling Policies - Zhu Jiekun, Quwan