Apache Cloudberry (Incubating) Support Tools

This repository contains a collection of tools to assist in supporting Apache Cloudberry (Incubating) deployments.

License

This project is licensed under the Apache License, Version 2.0. See LICENSE for the full license text.

Tools

Core File Analyzer

Location: scripts/apache_cloudberry_core_analyzer.py

A utility for analyzing PostgreSQL core files in Apache Cloudberry (Incubating) environments. It provides detailed analysis of core dumps and can identify patterns across multiple core files.

Features

• Single core file analysis with detailed GDB inspection
• Multiple core file analysis with pattern detection
• Signal and stack trace analysis
• Shared library dependency inspection
• JSON output for further processing
• Comparison of multiple core files to identify common crash patterns
• Custom core file pattern matching
• Pretty-printed JSON output for readability

Requirements

• Python 3.6+
• GDB (GNU Debugger)
• Apache Cloudberry (Incubating) environment (GPHOME must be set)
• Rocky Linux 8 compatible

Basic Usage

1. Analyze a single core file:

./apache_cloudberry_core_analyzer.py /path/to/core.1234

2. Analyze all core files in a directory:

./apache_cloudberry_core_analyzer.py /path/to/core/files/

3. Analyze and compare multiple core files:

./apache_cloudberry_core_analyzer.py /path/to/core/files/ --compare

4. Custom core pattern matching:

./apache_cloudberry_core_analyzer.py --core-pattern="core-*" /path/to/core/files/

5. View the JSON output:

# Using Python's json.tool
cat core_analysis_YYYYMMDD_HHMMSS.json | python3 -m json.tool

# Using less
less core_analysis_YYYYMMDD_HHMMSS.json

Output Format

The analyzer produces JSON output files for both individual analyses and comparisons:

• Individual analysis: core_analysis_YYYYMMDD_HHMMSS.json
• Comparison results: core_comparison_YYYYMMDD_HHMMSS.json

The JSON format is chosen for:

• Structured representation of analysis data
• Easy programmatic processing
• Human readability with proper formatting
• Tool integration capabilities

Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

Getting Help

For questions and support:

1. Check the documentation in the scripts/ directory
2. File an issue in the GitHub repository
3. Provide relevant sections of core analysis when reporting issues

Acknowledgments

This project builds on the excellent work of the PostgreSQL and GDB communities.