Data analytics with distributed tracing data
summary
Pipeline ingests Jaeger spans from Kafka, groups them in time windows, reconstructs traces into an in‑memory graph and runs analysis via notebook or streaming jobs deployed on OpenShift. Uses JupyterLab, Apache Spark and Kafka for batch and streaming, experiments with Cypher then Gremlin extensions plus TinkerGraph to simplify trace navigation, and exports derived metrics like service hop counts, trace depth and Prometheus‑exported trace quality metrics. Live demo shows extraction but not machine learning.
AI-written summary · about
chapters (11)
- 0:00Intro
- 2:12Goal
- 3:33Agenda
- 4:05Approach
- 7:33Tracing Data
- 9:13Graph Traversal
- 11:52Graph queries
- 12:39Trace metrics
- 15:07Prometheus metrics
- 17:44Open Source Metrics
- 18:50Questions
