Speed Up Java Debugging: One‑Click Tools to Pinpoint 100% CPU and Deadlocks

Background

Backend developers often encounter sudden high CPU load on production servers, especially during off‑hours, and need a fast, reliable method to locate the offending code.

Traditional 4‑step manual process

1. top -o ... with P:1040   // sort by process load
2. top -Hp <pid>          // find thread PID
3. printf "0x%x" <tid>   // convert to hex for jstack
4. jstack <pid> | vim +/0x... // view stack trace

This procedure is too cumbersome for urgent incident response.

One‑click tool: show-busy-java-threads.sh

The script automates the above steps, instantly identifying the most CPU‑intensive Java threads and displaying their stack traces.

source <(curl -fsSL https://raw.githubusercontent.com/oldratlee/useful-scripts/master/test-cases/self-installer.sh)

Example 1 – Regex backtracking causing 100% CPU

import java.util.*;
// ... Java code that matches complex regex patterns ...

Running the program creates a Java process that consumes 100% CPU. The tool pinpoints the thread executing Pattern$GroupHead.match in the regex engine.

Example 2 – Thread deadlock leading to hang

public class SimpleDeadLock extends Thread {
    static Object l1 = new Object();
    static Object l2 = new Object();
    // ... threads acquire locks in opposite order ...
}

The script reveals the circular lock acquisition, showing the exact lines where each thread waits for the other's lock.

Additional useful scripts

show-duplicate-java-classes : Detect duplicate classes in JARs or class directories.

find-in-jars : Search for classes or resources inside JAR files using regex.

housemd : JVM diagnostic tool supporting class loading, method tracing, and more.

jvm : Interactive menu for thread, GC, heap, and other JVM metrics.

greys : Non‑intrusive Java runtime tracing (Linux/Unix/Mac only).

sjk : Suite of JVM performance utilities (ttop, jps, hh, gc, etc.).

These tools together form a practical toolbox for quickly diagnosing CPU spikes, deadlocks, classpath conflicts, and other runtime issues in Java backend services.