Dates
Conferences
Tags
Sort by:
Most recent
Conference: KubeCon + CloudNativeCon North America 2022
Authors: Vineeth Pothulapati
2022-10-28
tldr - powered by Generative AI
Migrating from OpenTracing to OpenTelemetry for improved Tracer implementation and performance
- Two levels of migration: instrumentation layer and collector layer
- OpenTelemetry shim and complete re-instrumentation are two ways to migrate at the instrumentation layer
- Complete re-instrumentation offers more capabilities and flexibility
- OpenTelemetry SDK is more performant and has access to framework plugins
- Context propagation using baggage and headers is easy in OpenTelemetry
- OpenTelemetry supports multiple context propagations
- Demo shows how to migrate from OpenTracing to OpenTelemetry
Conference: KubeCon + CloudNativeCon Europe 2022
Authors: Reese Lee
2022-05-19
tldr - powered by Generative AI
The presentation discusses the importance of distributed tracing and the use of sampling strategies to manage the volume of data produced. It also highlights the challenges of implementing tail-based sampling using OpenTelemetry.
- Distributed tracing is important for understanding system connections and diagnosing problems.
- Traces are made up of spans, which represent logical units of work within a request.
- Sampling can be implemented at different stages of span processing to reduce the number of created or sampled spans.
- Tail-based sampling can be optimal for efficiently getting the desired data, but it can also have performance and scalability concerns.
- OpenTelemetry requires a collector to implement tail-based sampling, and all traces need to end up in the same collector for it to work properly.