Problem:
When running the start_server.sh script in a Docker container, the following errors occurred:
- : not foundserver.sh: 18: (and other lines)
- /app/start_server.sh: 35: Syntax error: "elif" unexpected (expecting "then")
Root cause:
The script file was using Windows-style line endings (CRLF) instead of Unix/Linux-style line endings (LF),
which caused syntax errors when executed in a Linux environment.
Solution:
1. Ensure start_server.sh file uses proper Unix line endings (LF)
2. Add dos2unix command in Dockerfile to convert line ending format of the script file
3. Automatically fix line ending issues during image build to ensure proper script execution in containers
This fixes the issue with starting Doris MCP Server in Linux-based Docker containers.
This PR refactors the configuration management to unify the handling of command-line arguments, environment variables, and default settings.
Key changes:
- All configuration is now consistently managed through the `DorisConfig` object.
- Command-line arguments now correctly override environment variables and default settings.
- The default server host/port is now correctly read from the configuration.
- Wrong environment variable loading in `schema_extractor.py` has been removed.
closes#37
1. **Unified Naming for CLI Arguments and Environment Variables**
- All database-related CLI arguments now use the `--doris-*` prefix, and environment variables use `DORIS_*` for consistency and maintainability.
- Backward compatibility: old `--db-*` arguments are still supported.
2. **Automatic Filtering of System SQL in Slow Query TopN**
- Slow query analysis now automatically excludes SQL statements involving `__internal_schema`, `information_schema`, and `mysql` system databases, ensuring only business-related slow queries are counted.
- Filtering is performed at the SQL level using `NOT LIKE` and `state != 'ERR'` for efficiency and safety.
3. **Unified Query Timeout Configuration**
- If no `timeout` is specified for query execution, the system will use the `config.performance.query_timeout` value as the default, falling back to 30 seconds if not configured.
- This avoids hardcoding and makes timeout management more flexible.
4. **Tool execution optimization**
- Significantly reduce the execution time of some data governance and operation and maintenance tools
- Optimize execution logic and reduce data scanning
- Enable concurrent scanning to speed up retrieval
5. **Log system optimization**
- Fix the Console log printing logic and output the log content correctly
- Add advanced tool execution process log output to facilitate further positioning of error locations
6. **DB Connection optimization**
- Fixed a connection pool acquisition exception caused by deadlock
7. **Other Improvements**
- Help documentation and CLI examples updated to reflect new and legacy parameter compatibility.
- Code comments and documentation further standardized for better team collaboration and open-source community understanding.
