Field guides for JVM operators
Heap dumps, thread dumps, GC logs, Native Memory Tracking, Flight Recorder: the evidence is built into the JDK. The Safepoint explains how to collect it, how to read it and which flags actually change the outcome, with every option checked against the official documentation.
Photo: Derrick Coetzee, Wikimedia Commons, CC0
Paste-able error messages, mapped to the guide that explains them
In focus · MemoryThe heap looks healthy, the process dies anyway. How NIO direct buffers are limited, why the default follows -Xmx, why -XX:+DisableExplicitGC makes things worse, and how to find who holds the buffers.
Heap, direct buffers, Metaspace, code cache and native memory: where a JVM's footprint goes and how to find a leak.
All memory guides
MemoryWhat counts as a GC root, the usual accidental references, and how to walk from a suspect object back to its root.
MemoryCapture safely with jcmd, dump automatically on OOM, then read the dominator tree and paths to GC roots.
MemoryHeap is only part of the process: read the NMT summary, diff against a baseline, and know what it cannot see.
MemoryWhat happens when compiled code runs out of room, the 240 MB default, segmented heaps and how to monitor them.
MemoryClass metadata outside the heap: the flags, the jcmd views, and how redeploys leak class loaders.
MemoryWhat AI demand did to DRAM prices, what a gigabyte costs in the cloud, and how to run the same JVM service on less memory.
Choosing a collector, tuning G1 and reading the unified GC log without guesswork.
All GC guides
Garbage collectionLatency, throughput or footprint: what each collector trades, and why your container may be running Serial.
Garbage collectionHeap size first, a realistic pause target, then IHOP and region size, and why -Xmn works against G1.
Garbage collectionOne line of configuration, the legacy flag mapping, and how to read pause lines for leaks and undersized heaps.
Threads & CPUTake a series, read states and lock lines, and let the JVM point out deadlocks for you.
Threads & CPUFrom top -H to the nid in a thread dump: a repeatable way to name the code that is burning CPU.
Threads & CPUCarriers, pinning before and after JDK 24, why not to pool them, and how to see them in dumps and JFR.
Running the JVM in containers, sizing the services around AI models and recording what the JVM does with Flight Recorder.
All production guides
ProductionHow limits are detected, what MaxRAMPercentage really sets, and a budget that keeps the OOM killer away.
ProductionContinuous recording, on-demand dumps with jcmd, the events that matter, and reading .jfr files.
ProductionBatching, quantisation and caching on the model side; heap, GC and container sizing for the Java services around it.
Each guide starts from a symptom seen in production and ends with commands you can run. Option names, defaults and behaviours are taken from primary sources: the java, jcmd and jfr reference pages, the HotSpot tuning and troubleshooting guides, JDK Enhancement Proposals and the OpenJDK source. Each guide lists them at the end.
Defaults change between JDK releases. Where a behaviour depends on the version, the guide says which release introduced or changed it. When in doubt, ask the JVM itself: java -XX:+PrintFlagsFinal -version prints every option with its effective value. More on the method on the about page; the full list is on all articles.
