Intelligent JVM Monitoring: Combining JDK Flight Recorder with AI
summary
Shows a working pipeline that streams JDK Flight Recorder events from each JVM via JFR streaming via Java agent into a central service, stores events, and runs a three‑agent LangChain4j workflow (data processor → anomaly detector → self‑healer) using Anthropic Haiku 4.5. The detector reports confidence rising from 72% to 90% after fetching older events, and the self‑healer invokes runtime changes over JMX self-healer actions to flip a batching flag successfully. Demo uses custom .jfc configs, agentic scopes for loops, and proves automated detection plus remediation with LangChain4j agentic workflow.
AI-written summary · about
chapters (13)
- 0:00Welcome
- 2:10Microservice architecture
- 3:17JFR monitoring challenges
- 5:06Centralized JFR streaming
- 10:29Java agent setup
- 12:55JMX fundamentals
- 15:47JMX for microservices
- 20:48AI agentic JFR analysis
- 25:25LangChain4j implementation
- 35:06Agent workflows in LangChain4j
- 44:20JMX runtime actions
- 48:17Live demo
- 49:00Summary
related talks
49mIntelligent JVM Monitoring: Combining JDK Flight Recorder with AI
42mContinuous Monitoring with JDK Flight Recorder
01h 09mContinuous Monitoring With JDK Flight Recorder (JFR)
46mContinuous Monitoring with JDK Flight Recorder
42mProfiling and monitoring with the JDK
39mProfiling and Monitoring with the JDK
18mContinuous Monitoring w/ JDK Flight Recorder (JFR)
45mAnalyzing Safe Flights, and Firey Wrecks with JDK Flight Recorder