Merge branch 'add-multi-worker-support' into max-graph-node

This commit is contained in:
yangdx
2025-03-02 18:37:56 +08:00

View File

@@ -102,8 +102,11 @@ light-server --llm-binding ollama --embedding-binding ollama
For production deployments, it's recommended to use Gunicorn as the WSGI server to handle concurrent requests efficiently. LightRAG provides a dedicated Gunicorn startup script that handles shared data initialization, process management, and other critical functionalities. For production deployments, it's recommended to use Gunicorn as the WSGI server to handle concurrent requests efficiently. LightRAG provides a dedicated Gunicorn startup script that handles shared data initialization, process management, and other critical functionalities.
```bash ```bash
# Start with run_with_gunicorn.py # Start with lightrag-gunicorn command
python run_with_gunicorn.py --workers 4 lightrag-gunicorn --workers 4
# Alternatively, you can use the module directly
python -m lightrag.api.run_with_gunicorn --workers 4
``` ```
The `--workers` parameter is crucial for performance: The `--workers` parameter is crucial for performance: