Skip to main content

Using AI thread dump analysis

A thread dump is vital data showing which threads are waiting on what inside the application. However, reading through hundreds of thread states to diagnose lock contention or blocking is heavy work even for experienced developers. WhaTap's AI thread dump analysis automates this so that operators without thread analysis experience can immediately identify problem threads and see improvement directions.

Prerequisites​

  • Project member or higher permission
  • The target instance is actively collecting data (visible on Instance Performance Management)

How to use it​

Path: Instance Performance Management > Thread list & dump > Thread dump button at the top right

  1. Select the target application.
  2. Click the Thread dump button to view the thread dump at that point in time.
  3. Click the Thread analysis button at the top of the dump.
  4. AI analysis starts (a few to tens of seconds).
  5. When analysis completes, the four-section result appears automatically in the drawer.

Managing analysis credits​

  • The top-bar Refresh button re-requests the analysis, which consumes a credit.
  • For a thread dump that has already been analyzed, clicking Thread analysis again does not consume a credit. The previous result is shown in the drawer as is.

Analysis result structure​

AI analyzes the collected thread dump and presents the result in four sections.

Table | AI analysis result structure
SectionContents
Overall summarySummary of the thread dump and total thread count analyzed
Thread state distributionPer-state thread status such as RUNNABLE, BLOCKED, and WAITING
Detected issuesLock contention, blocking, and wait issues that may impact performance
RecommendationsSuggested actions and recommended values for performance improvement

When to use it​

During incident response (most useful)​

Shorten thread dump interpretation time in step ③ Root cause analysis of the Incident response scenario.

  1. Select the suspect instance in Instance Performance Management.
  2. Select Thread list & dump, then click the Thread dump button.
  3. Click the Thread analysis button at the top of the thread dump.
  4. Start from the Detected issues AI surfaces, then drill down to the original thread.

Right after a release​

When response time spikes or threads pile up after deployment, capture a dump and run AI analysis to quickly distinguish whether the regression is lock contention or external resource wait.

Periodic performance tuning​

Collect dumps during normal operation and scan with AI to surface latent blocking and wait patterns humans miss, then feed them into the improvement backlog.

Understand the limits of AI analysis​

AI output is a first hypothesis, not the final conclusion. Use it this way:

  • For issues AI surfaces, cross-check against the original thread dump before acting.
  • Business context (recent deploys, external resource outages) is not visible to AI — combine with timeline info when interpreting.
  • Repeated analysis may not be fully consistent (LLM nature) — validate with observable metrics (thread count, response time).

Adjacent AI analysis features​

Table | Adjacent AI analysis features
FeatureWhen
AI active stack analysisIdentify bottlenecks in a transaction trace — AI active stack analysis
AI thread dump analysis (this guide)Interpret thread blocking and lock contention in an instance
AI SQL tuning guideIdentify inefficient queries and get execution-plan based improvements — complements DB realtime visibility
AI browser error stack analysisCode-level cause for frontend errors — complements RUM
WhaTap AI Chatbot / MCPNatural-language queries across WhaTap data — MCP integration

Next steps​

Last updated on