After implementing the B+ tree as the default index for the engine, I've encountered an issue where batch writes and reads exceed the expected runtime, despite the absence of apparent errors in checking. If anyone has insights or suggestions on potential causes or solutions for this performance issue, I would greatly appreciate your advice. Insights or recommendations on possible reasons or corrective measures for this performance anomaly would be highly valued. Expert advice is sought to aid in refining and optimizing the indexing implementation. Thank you in advance for any suggestions!

Phase 1: Basic

Develop

• Develop memory-disk architecture
• Data read/write process
• Design a database startup procedure
• Data deletion process

Test

• Unit tests
• DB test

Phase 2: Core Functionalities

• data file management strategy for efficient data write and retrieval
• LogRecord encoding/decoding
• Close and sync strategies
• Implement WriteBatch atomic writes

Phase 3: Performance Optimization

Develop

• #12
• Implement MMap for enhanced file I/O operations
• Establish data merge optimization strategies
• Develop HTTP APIs for database interaction over the web.

Test

• benchmark tests
• HTTP request tests
• Integrate Continuous Integration (CI) on GitHub Actions automate testing workflows
• Code coverage over 80%

Documentation

• Documentation 100%

Phase 4: Advanced Features and Integration

• Increased use of flat buffer option to support faster reading speed
• Integrate support for Redis data formats, enabling seamless migration and interoperability.

Roadmap Objectives and Milestones

• Q1-Q2: Complete architectural foundation and core functionalities setup.
• Q3: Focus on performance optimizations with advanced indexing and I/O strategies.
• Q4: Broaden database capabilities with advanced features.
• Q1 (Next): Extend Redis RESP protocol.

This roadmap is a living document and is subject to change as the project evolves. Welcome to write feedback or contributions to help shape the future of the project.