jstack Utility Deep Dive: JVM Thread Dumps, Deadlocks, and Mixed Stacks

Originally published on 2016-01-25; fully revised and updated.

The jstack utility is a JVM troubleshooting tool for capturing thread dumps and detecting deadlocks. This article is for DevOps and platform engineers who need fast, reliable thread-level diagnostics in production or pre-production environments, with practical commands and a clear triage workflow.

Where jstack fits in JVM diagnostics

Oracle JDK 8 troubleshooting documentation introduced Java Mission Control, Java Flight Recorder, and jcmd, and suggests using jcmd instead of jstack for enhanced diagnostics and lower overhead in many cases. Even so, jstack remains a direct and widely used way to obtain thread dumps, detect deadlocks, inspect core files, and print mixed Java/native stacks.

What jstack can collect

• Stack traces for all threads attached to a JVM, including Java threads and VM internal threads.
• Deadlock detection in Java-level locking scenarios.
• Ownable synchronizer details with -l (including java.util.concurrent.locks information).
• Mixed Java and native C/C++ frames with -m.
• Stack traces from a core dump.
• Remote querying via jsadebugd (with slower output, as noted in the source material).

Thread-dump methods and when to use them

Practical jstack command examples

Standard thread dump

jstack <pid>Code language: HTML, XML (xml)

Include ownable synchronizers

jstack -l <pid>Code language: HTML, XML (xml)

Force a stack dump from a hung process

jstack -F <pid>Code language: HTML, XML (xml)

Print mixed Java/native frames

jstack -m <pid>
jstack -m <pid> | c++filtCode language: HTML, XML (xml)

In mixed-stack output, frames prefixed with an asterisk (*) are Java frames; non-prefixed frames are native C/C++ frames.

Read stacks from a core dump

jstack $JAVA_HOME/bin/java coreCode language: PHP (php)

As documented by Oracle, you can also get equivalent thread-dump output via Ctrl+\ on the console or by sending QUIT to the JVM process.

Troubleshooting workflow for hangs and deadlocks

flowchart TD
    A[Service appears hung] --> B[Find JVM pid]
    B --> C[Run jstack pid]
    C --> D{Command responds}
    D -- yes --> E[Run jstack -l pid]
    D -- no --> F[Run jstack -F pid]
    E --> G[Check deadlock section]
    F --> G
    G --> H[If needed run jstack -m pid]
    H --> I[Identify lock or native stall]
    I --> J[Apply fix and validate]

How to read key output signals quickly

• Deadlock indicator: look for lines like Found one Java-level deadlock.
• Lock ownership clues: lines stating one thread is waiting to lock an object held by another thread.
• Thread state focus: prioritize BLOCKED, long WAITING, and repeating patterns across multiple dumps.
• Mixed stack clues: use -m when the issue may involve JNI or VM-native code paths.

References

• Troubleshoot with jstack Utility (Oracle JDK 8)
• Force a Stack Dump
• Stack Trace from a Core Dump
• Mixed Stack
• Control+Break Handler
• jstack command man page (Oracle JDK 8)

Best practices

• Start with jstack -l for lock-aware analysis when investigating hangs.
• If the process is unresponsive, use -F only when needed and note platform constraints.
• Use -m when Java-only stacks do not explain the issue (native/JNI suspicion).
• Keep a consistent incident runbook so thread dumps are captured quickly and repeatably.

Rajesh Kumar

I’m Rajesh Kumar, a DevOps, SRE, DevSecOps, Cloud, and Platform Engineering expert passionate about sharing practical knowledge, real-world experiences, and industry best practices. I have worked at Cotocus and regularly write about technology, travel, investing, health, product reviews, and digital marketing through my various platforms.

I publish technical articles at DevOps School, travel stories at Holiday Landmark, stock market insights at Stocks Mantra, health and fitness guidance at My Medic Plus, product reviews at TrueReviewNow, and SEO and digital marketing strategies at Wizbrand.