Post-Mortem JVM Crash Analysis with jcmd
summary
Adds JEP 528 to let jcmd run against core/minidump files by launching a second revival helper process that mmap’s the crashed process image, loads the same libjvm, and invokes the JVM diagnostic DCMD parse/execute path to produce thread dumps, heap dumps, class histograms, metaspace and code‑cache reports. Works on Linux x64 and aarch64 and Windows x64 today, requires the original JVM binary (-L) and explicit core file option (-C), and cannot run live actions like GC.run or JFR start; targeting JDK 27 integration.
AI-written summary · about
chapters (9)
- 0:00Post-mortem JVM crash analysis intro
- 3:57Why standard tools fail
- 8:47Native debugger limits
- 12:17Serviceability Agent challenges
- 16:51Extending jcmd for core files
- 22:27Revival process restores state
- 28:09jcmd core analysis limits
- 33:43JEP 528 status and plans
- 43:05Q&A on memory mapping
related talks
50mEmergency Dump — The Black Box of a Crashed JVM
53mServiceability in J9
31mJava Mission Control: To Infinity and Beyond…
46mAnalyzing HotSpot Crashes
40mA Glance At The Java Performance Toolbox
44mA Glance At The Java Performance Toolbox
47mLegal JVM Doping in Your Containers
45mUnder the Hood of a JVM Application - Java Flight Recorder / Java Mission Control