debug-hang

Name: debug-hang
Availability: InStock
Author: enkez

Diagnose Ray distributed training hang issues by analyzing cluster status and blocking patterns.

1|Updated Jun 3, 2026

Search Tags:#distributed training #Ray #cluster analysis #performance troubleshooting #hang diagnosis

Ready to run

npx skills add https://github.com/EnkeZ/tgsopd --skill debug-hang-enkez

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: debug-hang
Source: https://github.com/EnkeZ/tgsopd/tree/main/skills/debug-hang
Command: npx skills add https://github.com/EnkeZ/tgsopd --skill debug-hang-enkez

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires ray, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill automates the diagnosis of Ray distributed training hang issues, providing a streamlined process to identify and resolve problems with Ray scheduling tasks.

Core Features & Use Cases

Cluster Status Overview: Provides a summary of the cluster's health and resource usage.
Task Blocking Identification: Locates the tasks that are causing the hang.
Call Stack Collection: Gathers the call stacks for the running tasks.
Actor Dependency Analysis: Analyzes the dependencies of the actors.
Blocking Pattern Matching: Matches the blocking patterns to potential causes.
Root Cause Identification: Determines the root cause of the hang.
Automated Diagnosis Script: Offers a script to automate the diagnosis process.
Case Studies: Provides reference cases for similar hang issues.

Quick Start

Run the diagnosis script to automatically troubleshoot Ray distributed training hang issues.

Frequently Asked Questions about debug-hang

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I diagnose Ray distributed training hang issues?▼

Diagnosing Ray distributed training hang issues involves analyzing cluster status, task blocking, call stacks, and actor dependencies. An automated diagnosis script matches blocking patterns to determine the root cause of the hang.

What causes tasks to block and hang during Ray distributed training?▼

Tasks block and hang during Ray distributed training often due to actor dependency deadlocks or resource exhaustion. Analyzing call stacks and matching blocking patterns pinpoints the exact root cause.

Do I need Ray pre-installed to troubleshoot cluster hang problems?▼

Yes, you need Ray and Python scripts installed to troubleshoot cluster hang problems. The diagnosis process relies on these dependencies to collect call stacks and analyze actor dependencies.

Can I analyze actor dependencies to resolve Ray scheduling hangs?▼

Yes, you can analyze actor dependencies to resolve Ray scheduling hangs. The diagnosis process includes actor dependency analysis and blocking pattern matching to locate the tasks causing the hang.

What is the best way to troubleshoot Ray hangs in large-scale distributed training environments?▼

The best way to troubleshoot Ray hangs in large-scale distributed training environments is using an automated diagnosis script that collects call stacks, evaluates cluster health, and matches blocking patterns.