BIZ-46 Phase3: 7项 follow-up 开发完成

1. 架构解耦 — SidecarContext + FastAPI Depends 注入
   - 新增 context.py: SidecarContext dataclass 收敛全部全局状态
   - server.py: 移除模块级全局变量，lifespan 创建 ctx → app.state.sidecar
   - webui.py: 移除反向导入 server，改用 Depends(get_context)

2. Prometheus 标签基数治理 — model_id → provider
   - upstream_latency_seconds / upstream_errors_total label 收敛为 provider
   - 模型级信息保留在 structlog JSON 日志

3. SSE 快照共享缓存
   - 1s TTL 共享 snapshot cache + double-check locking
   - 多客户端不重复构建快照

4. 部署支撑
   - Dockerfile (python:3.12-slim, 非 root 用户, HEALTHCHECK)
   - systemd service (安全加固, 资源限制)
   - .env.example (完整环境变量清单)

5. Readiness HTTP Client 复用
   - check_upstream() 注入主 http_client，不再每次创建新 client

6. Retreat 并发回归测试
   - 5 个测试用例全部通过（死锁检测 + 状态转换 + 并发安全）

7. Dashboard UX 优化
   - 队列柱状图 300ms 平滑动画
   - SSE 断连 5s 半透明遮罩
   - 队列图标题显示总排队数
   - 页面加载同步配置

验证: mypy strict 通过 (0 errors), pytest 5/5 通过, server 导入正常 (13 routes)

Co-authored-by: multica-agent <github@multica.ai>

services/nvidia_sidecar/context.py

@@ -0,0 +1,75 @@
+"""
+NVIDIA Sidecar — SidecarContext 依赖注入容器 (§BIZ-46 Phase3)
+
+将所有模块级全局状态收敛为单一 dataclass，通过 FastAPI app.state 注入，
+消除 webui.py → server 的反向导入，支持可测试性和多实例扩展。
+
+设计文档: docs/architecture/BIZ-46_Phase3_Architecture_Design.md §1
+"""
+
+from __future__ import annotations
+
+import asyncio
+import time
+from dataclasses import dataclass, field
+from typing import TYPE_CHECKING, Any
+
+import httpx
+
+if TYPE_CHECKING:
+    from nvidia_sidecar.config import SidecarConfig
+    from nvidia_sidecar.rate_limiter import AdaptiveTokenBucket
+    from nvidia_sidecar.priority_queue import PriorityRequestQueue
+    from nvidia_sidecar.metrics import PrometheusMetrics
+    from nvidia_sidecar.health import HealthService
+
+
+@dataclass
+class SidecarContext:
+    """Sidecar 全局运行时上下文 — 所有核心组件的唯一容器。
+
+    通过 ``app.state.sidecar`` 注入 FastAPI，路由通过 ``Depends(get_context)`` 获取。
+    """
+
+    # ---- 核心组件 ----
+    config: SidecarConfig
+    http_client: httpx.AsyncClient
+    token_bucket: AdaptiveTokenBucket
+    priority_queue: PriorityRequestQueue
+    prometheus: PrometheusMetrics
+    health: HealthService
+
+    # ---- 运行时状态 ----
+    pending_requests: dict[str, tuple["asyncio.Future[Any]", float]] = field(default_factory=dict)
+    """request_id → (response future, enqueued_at) 的映射。"""
+
+    stats: dict[str, int] = field(default_factory=lambda: {
+        "total_requests": 0,
+        "nvidia_requests": 0,
+        "passthrough_requests": 0,
+        "ratelimited_requests": 0,
+        "queue_full_rejects": 0,
+        "upstream_errors": 0,
+        "start_time": 0,
+    })
+
+    stats_lock: asyncio.Lock = field(default_factory=asyncio.Lock)
+
+    # ---- 缓存 ----
+    snapshot_cache: tuple["dict[str, Any]", float] | None = None
+    """SSE 快照共享缓存: (data, timestamp)。"""
+    snapshot_cache_lock: asyncio.Lock = field(default_factory=asyncio.Lock)
+    SNAPSHOT_CACHE_TTL: float = 1.0
+
+    # ---- 便捷方法 ----
+
+    async def increment_stat(self, key: str, delta: int = 1) -> None:
+        """线程安全的统计计数器自增。"""
+        async with self.stats_lock:
+            self.stats[key] = self.stats.get(key, 0) + delta
+
+    @property
+    def uptime_seconds(self) -> int:
+        """服务运行时长（秒）。"""
+        st = self.stats.get("start_time", 0)
+        return int(time.time() - st) if st else 0