Why My Java App Hits 100% CPU: Live Infinite Loop Demo & Diagnosis

CPU usage hitting 100% can have many causes; an infinite loop in code is a classic one. To explore this, I created a controlled experiment (not in production) using Vagrant, VirtualBox, and Ansible to provision a simple Spring MVC application that contains an endpoint with an intentional infinite loop.

The loop is exposed at http://192.168.88.10:9898/web/loop. After hitting the URL twice, the CPU stays around 120‑130% instead of 200%.

1. Using top to locate the offending process

First I opened a terminal and ran top to see which process was consuming CPU. The request URL was accessed, and the top output showed the Java process with elevated CPU usage.

2. Checking JVM heap with jstat

The issue does not involve heap pressure, so I ran jstat -gcutil 32593 1s (where 32593 is the Java PID) to sample heap usage every second. The output showed normal heap utilization.

3. Investigating the stack to find the busy thread

Since the heap was fine, I listed the Java threads with top -H -p <java_pid> to see which thread was using CPU.

Next I dumped the stack of the suspect thread using jstack -l <thread_pid> >> stack.log (chosen thread PID 3596, which is 0xe0c in hexadecimal).

4. Summary

Through this step‑by‑step investigation—using top to locate the process, jstat to rule out heap issues, and jstack to pinpoint the looping thread—we demonstrated a practical method for diagnosing a Java application that drives CPU usage to 100%.