Files
furyhawk bb4b614711 feat: Implement RAG document and sync services with vector store integration
- Added `vectorstore.py` for managing vector storage operations using Milvus.
- Introduced `rag_document.py` for handling RAG document lifecycle, including ingestion and status updates.
- Created `rag_status.py` for streaming RAG ingestion status via Redis pub/sub.
- Developed `rag_sync.py` for managing synchronization operations and logs.
- Implemented `sync_source.py` for managing sync source configurations and triggering syncs.
- Established a lightweight task dispatcher in `dispatcher.py` for background task execution.
- Registered task functions for RAG ingestion and sync operations in `rag_tasks.py`.
- Configured task dispatcher settings in `arq_settings.py`.
- Added necessary database models and schemas for RAG document and sync operations.
2026-06-13 21:52:27 +08:00

27 lines
738 B
Python

"""Task dispatcher configuration.
In the current implementation, tasks run in-process via
:class:`backend.worker.dispatcher.TaskDispatcher`. For production
deployments with multiple replicas, replace with a Redis-backed
queue (ARQ, Taskiq, or Celery).
To switch to ARQ in the future::
pip install arq
arq backend.worker.arq_settings.WorkerSettings
"""
from __future__ import annotations
from backend.core.config import settings
from backend.worker.tasks.rag_tasks import TASK_REGISTRY
# List of registered task functions for ARQ/Taskiq configuration
FUNCTIONS = list(TASK_REGISTRY.values())
# Redis connection info (for future ARQ/Taskiq worker setup)
REDIS_URL = settings.valkey_url
__all__ = ["FUNCTIONS", "REDIS_URL"]