Jaeger Deep Dive

Overview

Dive deep into the Jaeger project with this comprehensive conference talk featuring experts from Facebook, Red Hat, and Grafana Labs. Explore recently added features through a short demo, and gain insights into Jaeger's architecture, deployment models, configuration options, sampling techniques, and multi-tenancy capabilities. Learn about integration with OpenTelemetry, best practices, and the project's roadmap. Discover how to effectively deploy Jaeger, maximize its benefits, and contribute to the project. Topics covered include metrics, context propagation, distributed tracing, service dependencies, trace timelines, sampling techniques, instrumentation, and deployment on Kubernetes.

Syllabus

Intro
Metrics - Cardinality
Context Propagation & Distributed Tracing
Service dependencies diagram
Transitive Service Graphs
Trace timeline - A blocking operation
Trace timeline - Sequential operations
Comparing trace structures - Unified diff
Jaeger, a Distributed Tracing Platform
Instrumentation not included
Jaeger and your system
Architecture 2017: Push
Technology Stack
Why do we sample
Sampling techniques
Head-based sampling in Jaeger
Jaeger sampling configuration
Tail-based (post-trace) sampling
Tail-based sampling in Jaeger
Jaeger with OpenTracing
Jaeger with OpenTelemetry
Deploying Jaeger on Kubernetes