feat: 异步任务队列架构 - SQLite队列 + 速率限制 + NAS Poller

Edge端:
- 新增 SQLite 异步任务队列 (queue_manager + consumer)
- 新增 TokenBucket 速率限制器 (Gemini 1000 RPM, NVIDIA 40 RPM, burst 2x)
- 新增 /api/edge/video/enqueue + /api/edge/results 端点
- 消费者线程从队列消费任务，按速率限制调用AI模型
- orchestrator 集成 rate_limiter，Gemini优先→NVIDIA兜底

NAS端:
- Dispatcher 重构为 enqueue 模式（上传后立即返回，不等结果）
- 新增 Poller 线程（定期从Edge拉取结果写 MariaDB）
- app.py 启动 Poller，config.yaml 新增 poller 配置
- db_layer 更新 valid_stages 添加 'process'

fam-edge/src/fam_edge/ai_orchestrator/orchestrator.py

@@ -52,12 +52,15 @@ class AIOrchestrator:
     def run_visual_analysis(self, adapters: List[BaseModelAdapter],
                            frame_paths: List[str],
                            frame_timestamps: List[str],
-                           known_members_context: str) -> Dict[str, dict]:
+                           known_members_context: str,
+                           rate_limiter=None) -> Dict[str, dict]:
         """视觉分析阶段：仅 role=vision 的适配器参与
 
         orchestrator.mode:
         - fallback (默认): 按 config 顺序依次尝试，首个成功即采用（单元素 dict）
         - ensemble: 并行所有健康 vision 模型，全部成功结果都保留（交叉验证）
+
+        rate_limiter: 可选 RateLimiter 实例，按 provider 限速（2x burst）
         """
         vision_adapters = [a for a in adapters if getattr(a, 'role', 'vision') == 'vision']
         if not vision_adapters:
@@ -68,7 +71,8 @@ class AIOrchestrator:
 
         if mode == 'ensemble':
             return self._run_visual_ensemble(
-                vision_adapters, frame_paths, frame_timestamps, known_members_context)
+                vision_adapters, frame_paths, frame_timestamps,
+                known_members_context, rate_limiter)
 
         # fallback: 顺序降级，首个成功即采用
         model_outputs = {}
@@ -76,6 +80,12 @@ class AIOrchestrator:
             if adapter.get_circuit_breaker().is_open():
                 logger.warning(f"[{adapter.provider_name}] 熔断器 OPEN，跳过")
                 continue
+            # 速率限制：按 provider 获取 token（2x burst）
+            if rate_limiter:
+                acquired = rate_limiter.acquire(adapter.provider_name, timeout=300)
+                if not acquired:
+                    logger.warning(f"[{adapter.provider_name}] 速率限制超时，跳过")
+                    continue
             start = time.time()
             try:
                 output = adapter.analyze_frames(
@@ -97,7 +107,8 @@ class AIOrchestrator:
         return model_outputs
 
     def _run_visual_ensemble(self, vision_adapters, frame_paths,
-                             frame_timestamps, known_members_context) -> Dict[str, dict]:
+                             frame_timestamps, known_members_context,
+                             rate_limiter=None) -> Dict[str, dict]:
         """并行调用所有健康 vision 模型，保留全部成功结果（交叉验证）"""
         model_outputs = {}
         max_timeout = max((a.get_timeout() for a in vision_adapters), default=240)
@@ -107,6 +118,12 @@ class AIOrchestrator:
                 if adapter.get_circuit_breaker().is_open():
                     logger.warning(f"[{adapter.provider_name}] 熔断器 OPEN，跳过")
                     continue
+                # 速率限制：按 provider 获取 token（2x burst）
+                if rate_limiter:
+                    acquired = rate_limiter.acquire(adapter.provider_name, timeout=300)
+                    if not acquired:
+                        logger.warning(f"[{adapter.provider_name}] 速率限制超时，跳过")
+                        continue
                 future = pool.submit(
                     adapter.analyze_frames,
                     frame_paths, frame_timestamps, known_members_context)
@@ -377,9 +394,12 @@ class AIOrchestrator:
         return 200
 
     def process_push_task(self, task_data: dict, video_path: str,
-                          preprocessor: 'VideoPreprocessor') -> dict:
+                          preprocessor: 'VideoPreprocessor',
+                          rate_limiter=None) -> dict:
         """推送模式：同步处理上传的视频，结果直接返回（无 webhook 回调）
 
+        rate_limiter: 可选 RateLimiter 实例，按 provider 限速（2x burst）
+
         返回 payload 结构与原 webhook 回调一致:
         - 成功: {task_id, status: "success", event_start_time, ..., frame_details, ...}
         - 失败: {task_id, status: "failed", failure_stage, error_message}
@@ -429,7 +449,8 @@ class AIOrchestrator:
 
             # 3. 并行视觉分析
             model_outputs = self.run_visual_analysis(
-                healthy_adapters, compressed_frames, frame_timestamps, known_members
+                healthy_adapters, compressed_frames, frame_timestamps,
+                known_members, rate_limiter
             )
             if not model_outputs:
                 raise Exception('All models failed in visual analysis')