The art and methodology of troubleshooting
summary
Starts with concrete triage: monitor first, profile later using Zabbix/NewRelic/Kibana to spot memory leaks and latency spikes before touching code. Shows real fixes: replacing SecureRandom reduced thread blocking, disabling MySQL server-side prepared statements closed a leak that built up over two days, fixing a Disruptor ring-buffer cut surviving young-gen from ~700MB to ~200MB, and Java 8 halved CPU and GC pause times. Demonstrates using heap dumps + Eclipse MAT and Java Flight Recorder to find root causes.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 2:00Monitoring first step
- 8:50Stack traces and locks analysis
- 13:09Memory dumps usage
- 20:11Java Flight Recorder
- 25:12Code and allocation profiling
- 31:18Content buses diagnostics
- 44:43Handling errors and exceptions
- 50:19Garbage collection monitoring
- 1:00:22Q&A
related talks
38mLost Art of Troubleshooting
39mHow to Use Common Investigation Practices in Troubleshooting
50mDebugging complex issues in web applications
46mProfiling your Java Application A Beginner’s Guide
28mArt of Debugging
33mLove your bugs
34mMonitoring for devs. The necessity or self-indulgence
43mSoftware Operations and Human Cognition