Master JVM Performance: Using jps, jstack, jmap, jstat & hprof

JVM Process Status (jps)

The jps tool lists Java processes and their main class or JAR. Basic syntax is jps [options] [hostid]. Common options include -q (quiet), -m (show main arguments), -l (full class name), and -v (JVM arguments). Example output shows process IDs and the corresponding main classes. jps -m -l Sample output on a server:

2458 org.artifactory.standalone.main.Main /usr/local/artifactory-2.2.5/etc/jetty.xml
29920 com.sun.tools.hat.Main -port 9998 /tmp/dump.dat
3149 org.apache.catalina.startup.Bootstrap start
...

Thread Stack Inspection (jstack)

The jstack utility prints stack traces of all threads in a Java process, useful for locating deadlocks or high‑CPU threads. Syntax: jstack [option] pid. Options such as -l add lock information. jstack -l 21711 | grep 54ee In the example, the most CPU‑intensive thread (ID 21742) is identified, converted to hexadecimal (0x54ee), and then filtered from the stack dump to reveal it is waiting in PollIntervalRetrySchedulerThread.

"PollIntervalRetrySchedulerThread" prio=10 tid=0x00007f950043e000 nid=0x54ee in Object.wait() [0x00007f94c6eda000]

The corresponding Java source shows an idle wait using sigLock.wait(timeUntilContinue), confirming the hotspot.

// Idle wait
getLog().info("Thread [" + getName() + "] is idle waiting...");
schedulerThreadState = PollTaskSchedulerThreadState.IdleWaiting;
long now = System.currentTimeMillis();
long waitTime = now + getIdleWaitTime();
long timeUntilContinue = waitTime - now;
synchronized(sigLock) {
    if (!halted.get()) {
        sigLock.wait(timeUntilContinue);
    }
}

Heap Inspection (jmap) and Heap Analysis (jhat)

The jmap tool inspects heap memory. Common usages include: jmap -heap pid – shows heap configuration and usage. jmap -histo[:live] pid – prints a histogram of object counts and sizes. jmap -dump:format=b,file=heap.bin pid – creates a binary heap dump.

Example of jmap -heap output (truncated):

Heap Configuration:
MinHeapFreeRatio = 40
MaxHeapFreeRatio = 70
MaxHeapSize = 2067791872 (1972.0MB)
...
Heap Usage:
PS Young Generation
Eden Space: capacity = 6422528 (6.125MB), used = 5445552 (5.19MB)
...

After generating a heap dump, jhat can be used to explore it via a web UI: jhat -port 9998 /tmp/dump.dat The UI allows OQL queries to locate objects of interest.

JVM Statistics (jstat)

jstat

reports JVM statistics such as garbage‑collection metrics. Syntax: jstat [options] vmid [interval] [count]. Example showing GC statistics every 250 ms for four samples:

jstat -gc 21711 250 4

S0C    S1C    S0U    S1U    EC      EU      OC       OU      PC    PU   YGC  YGCT  FGC  FGCT   GCT
192.0 192.0  64.0   0.0  6144.0 1854.9  32000.0 4111.6 55296.0 25472.7 702 0.431 3 0.218 0.649

Key columns:

S0C/S1C – Survivor space capacities; S0U/S1U – used.

EC/EU – Eden space capacity/used.

OC/OU – Old generation capacity/used.

PC/PU – Permanent generation (or Metaspace) capacity/used.

YGC/YGCT – Young GC count and time.

FGC/FGCT – Full GC count and time.

GCT – Total GC time.

Heap/CPU Profiling (hprof)

The legacy hprof agent can profile CPU usage, heap allocation, or generate heap dumps. It is invoked via JVM arguments, e.g.:

java -agentlib:hprof=cpu=samples,interval=20,depth=3 MyApp

Options include: cpu=samples|times – sampling or exact method‑level timing. heap=sites|dump|all – allocation sites or full dump. format=a|b – text or binary output.

Various tuning parameters such as interval, depth, and file.

Example of CPU sampling profile generation:

java -agentlib:hprof=cpu=samples,interval=20,depth=3 Hello

For detailed heap analysis, heap=dump produces a binary dump that can be opened with tools like MAT or VisualVM.

Note: Enabling -Xrunprof:heap=sites on production servers can significantly impact performance and is not recommended.