perf: 优化请求鉴权链路并批量化统计/调度查询

- 为 Pipeline/Context 增加按需读取请求体能力,支持 async 懒加载 JSON body
- 为 chat/cli/video/claude/openai-cli 适配器关闭默认预读,减少无效 body 读取与超时风险
- 将本地登录、JWT 用户加载、API Key 鉴权迁移到线程池隔离会话执行,避免阻塞事件循环
- 为 API Key 鉴权返回结构化余额结果,并在主请求会话中重新绑定 user/api_key 后再校验状态、过期和锁定信息
- 为 management/user token 前缀认证引入独立会话与结果回绑,避免跨会话对象写入失效

- 为 Usage 余额检查补充结构化返回,统一透出 remaining 与欠费/不可用文案映射
- 为用户与管理端活跃请求查询增加 maintain_status 开关,避免轮询指定 id 时误触发状态修复
- 重写 user_me usage 汇总逻辑,支持 group_by=None 的粗粒度聚合
- 修正 provider 维度成功率与平均响应时间统计,基于 success_count 和成功响应耗时汇总计算
- 前端 Usage 轮询由 setInterval 改为串行 setTimeout,避免并发轮询叠加

- 为 StatsAggregator 增加按本地日期批量计算百分位能力,替代逐天 fan-out 查询
- 为混合统计查询合并连续实时日期区间,并批量读取 StatsDaily,减少逐日查询次数
- 为用户日统计增加批量聚合入口,替代逐用户循环聚合
- 为系统配置导出改用 selectinload 预加载 provider 关联数据,减少 N+1 查询
- 为管理员用户列表增加钱包批量查询,避免逐用户回表

- 为调度器增加 provider 轻量引用预过滤,先按 allowed_providers 缩小范围再加载完整 provider 图
- 为 CandidateBuilder 增加 provider refs/provider_ids 查询能力,保留分页顺序
- 为模型缓存增加 provider_model_mappings 索引缓存与 model_mappings 规则缓存,减少重复全量扫描
- 为请求候选中间态改为 flush/batch commit,降低 pending/streaming 状态切换的事务往返
- 为钱包访问结果补充 balance_snapshot,并抽取余额快照复用逻辑

- 补充 pipeline、auth、admin users、user_me usage、stats aggregator、model cache、
  scheduler、wallet、request candidate 等回归与契约测试
This commit is contained in:
AAEE86
2026-03-09 22:57:23 +08:00
parent bf818e3b61
commit 57c7cca556
37 changed files with 2490 additions and 498 deletions

View File

@@ -11,6 +11,13 @@ from src.api.base.context import ApiRequestContext
def _build_request(headers: dict[str, str] | None = None) -> Request:
return _build_request_with_body(b"", headers=headers)
def _build_request_with_body(
body: bytes,
headers: dict[str, str] | None = None,
) -> Request:
header_items = [
(str(key).encode("latin-1"), str(value).encode("latin-1"))
for key, value in (headers or {}).items()
@@ -28,8 +35,14 @@ def _build_request(headers: dict[str, str] | None = None) -> Request:
"server": ("testserver", 80),
}
received = False
async def receive() -> dict[str, object]:
return {"type": "http.request", "body": b"", "more_body": False}
nonlocal received
if received:
return {"type": "http.request", "body": b"", "more_body": False}
received = True
return {"type": "http.request", "body": body, "more_body": False}
request = Request(scope, receive)
request.state.perf_metrics = {}
@@ -89,3 +102,22 @@ class TestApiRequestContextEnsureJsonBody:
assert context.client_content_encoding == "gzip"
assert context.client_accept_encoding == "gzip, deflate"
@pytest.mark.asyncio
async def test_ensure_json_body_async_loads_body_lazily(self) -> None:
payload = {"message": "hello", "count": 2}
request = _build_request_with_body(json.dumps(payload).encode("utf-8"))
context = ApiRequestContext.build(
request=request,
db=None, # type: ignore[arg-type]
user=None,
api_key=None,
raw_body=None,
)
assert context.raw_body is None
result = await context.ensure_json_body_async()
assert result == payload
assert context.raw_body == json.dumps(payload).encode("utf-8")

View File

@@ -0,0 +1,40 @@
from __future__ import annotations
from types import SimpleNamespace
from unittest.mock import MagicMock
from src.services.request.candidate import RequestCandidateService
def _build_db_with_candidate(candidate: SimpleNamespace) -> MagicMock:
query = MagicMock()
query.filter.return_value.first.return_value = candidate
db = MagicMock()
db.query.return_value = query
db.info = {"managed_by_middleware": True}
return db
def test_mark_candidate_started_flushes_without_immediate_commit() -> None:
candidate = SimpleNamespace(status="available", started_at=None)
db = _build_db_with_candidate(candidate)
RequestCandidateService.mark_candidate_started(db, "candidate-1")
assert candidate.status == "pending"
assert candidate.started_at is not None
db.flush.assert_called_once()
db.commit.assert_not_called()
def test_mark_candidate_streaming_flushes_without_immediate_commit() -> None:
candidate = SimpleNamespace(status="pending", concurrent_requests=None)
db = _build_db_with_candidate(candidate)
RequestCandidateService.mark_candidate_streaming(db, "candidate-1", concurrent_requests=3)
assert candidate.status == "streaming"
assert candidate.concurrent_requests == 3
db.flush.assert_called_once()
db.commit.assert_not_called()