Monitoring Reactive Applications
summary
Lightbend builds a monitoring agent that instruments Akka/Play/Scala apps, captures metrics and traces in the JVM instead of emitting millions of offline events, and produces histograms and trace context IDs so asynchronous paths can be reassembled. It supports pull (Prometheus) and push models, sampling and delta strategies to limit overhead, runtime dynamic configuration to raise monitoring on hot paths, dispatcher/thread-pool visibility to spot blocked actors, and roadmap items like Futures/Streams flows and Zipkin-style distributed tracing.
AI-written summary · about
chapters (11)
- 0:00Intro
- 0:52What is Monitoring?
- 3:45Traditional Monitoring: Metrics and Debugging
- 6:13Architecture Shift: Cloud Computing
- 12:23Monitoring Reactive Applications: The Challenge
- 15:30Monitoring Reactive Applications: Tracing
- 16:35Monitoring Reactive Applications: Dapper
- 19:36Monitoring Reactive Applications: Trace Events
- 26:56Trends in Monitoring: Pull/Push - System Status
- 32:50Trends in Monitoring: Dynamic Configuration
- 37:22Best Practice
related talks
52mIs your JVM app flying blind? Unmask hidden issues with Observability superpowers!
48mSetting up monitoring for Java apps
52mBuilding Reactive Apps
01h 04mTracing distributed services: experiences with implementing APM for the JVM
42mDistributed Monitoring: How to understand the Chaos
43mDistributed Monitoring: How to Understand the Chaos
01h 10mObservability For (Java) Devs - 2023 Edition
45mLatency tracing in distributed Java applications