Understanding Microservices with Distributed Tracing
summary
Shows that instrumenting microservices with OpenTracing and vendor tools like Jaeger or LightStep yields concrete wins: zero‑touch collection via an Envoy sidecar rolled out once across hosts produces usable traces across 900 services, where single requests often generate 200+ RPC calls. Traces expose the exact span taking latency or repeated calls (leading to caching fixes), integrate with dashboards and PagerDuty, and surface sampling and naming as practical challenges.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 0:29Background
- 2:19Agenda
- 3:35Observability Tools
- 5:23Metrics and Logging
- 8:54Query language
- 14:17Adding instrumentation
- 17:55Open Tracing
- 21:50Envoy
- 24:37Challenges
related talks
51mDistributed Tracing for microservices with Jaeger
38mDistributed Tracing: Practices, Patterns & Technologies for Debugging Microservices
40mCracking Microservices Performance Issues with Distributed Tracing
34mWould You Like Some Tracing With Your Monitoring?
43mDebugging and Profiling Distributed Applications Using OpenTracing
34mReduce Complexity with Distributing Tracing
16mThe State of Observability
47mUsing metric and traces to diagnose our cloud-native applications