How do you use Java thread dumps to diagnose a slow or stuck service?

Uses multiple thread dumps, CPU correlation, lock ownership, executor context, and request evidence to distinguish deadlock, contention, blocking, and hot loops.

1Say itAnswer out loud first

2CompareCompare with Core answer

3RefineRefine with Stronger and Advanced

TL;DR

I capture several thread dumps a few seconds apart and correlate them with CPU, request latency, executor metrics, and the incident timeline. Repeated identical stacks can show a stable lock wait, blocked I/O, or a hot loop, while changing stacks may show slow progress. A dump identifies thread state, stack frames, owned monitors, and locks being awaited. JVM-reported deadlock information is useful, but not every stuck service is a formal deadlock.

Full answer below

Practice focus

Production evidence collection and interpretation rather than reading one isolated stack trace.

Answer first

Answer path

Core answer first, then checkpoints toward a stronger answer.

Start
Core answer
Start here. This is the direct spoken answer to practice first.
Overview
A thread dump is a point-in-time view of thread states and stacks; several samples reveal whether work is progressing.
I capture several thread dumps a few seconds apart and correlate them with CPU, request latency, executor metrics, and the incident timeline. Repeated identical stacks can show a stable lock wait, blocked I/O, or a hot loop, while changing stacks may show slow progress. A dump identifies thread state, stack frames, owned monitors, and locks being awaited. JVM-reported deadlock information is useful, but not every stuck service is a formal deadlock.
Checkpoint
Stronger answer
Optional upgrade with examples, trade-offs, and sharper phrasing.
For high CPU, I correlate operating-system thread CPU with the JVM thread identifier and inspect repeated hot stacks. For low CPU and high latency, I look for workers waiting on downstream I/O, exhausted connection pools, locks, futures, or executor queues. Thread names and executor-specific naming make ownership much easier to identify. One dump during normal activity can look alarming, so comparison and workload context prevent false conclusions.

Review signals

Use these to catch weak phrasing, shallow reasoning, and risky claims.

Common mistakesCandidateCommon mistakesMistakes that weaken the answer or signal shallow understanding.Open details

Details

Drawing a conclusion from one dump.
Calling every WAITING thread a problem.
Restarting before preserving evidence.

How do you use Java thread dumps to diagnose a slow or stuck service?

Practice focus

Answer path

Overview

Review signals

Follow-up prompts

Continue practice

Skill areas

Suggest an improvement

Found this answer useful?