feat(proxy): T-P2 上游客户端(流式派发/usage 注入过滤/三家归一化/首 token 前 failover)

- upstream.py:httpx.AsyncClient 模块级单例(keepalive,limits=100,
  超时 connect10/read120/write10/pool30);stream() 始终注入
  stream_options.include_usage(计量不依赖客户端)+ filter_usage_chunk
  (客户端未要求 usage 时剥除该 chunk);D-P4 首 token 前 failover 链、
  流中失败抛 UpstreamAborted(不可切换);ttfb_ms 记录
- normalize_usage 三家归一:deepseek prompt_cache_hit_tokens /
  openai prompt_tokens_details.cached_tokens / anthropic cache_read_input_tokens
  (命中数>总数时钳制)
- model_pool:+provider(枚举校验 deepseek/openai/anthropic)+in_hit_price
  (缺省 = price_in×1/30,D-P3)
- 测试 +6:透传+sink/failover/流中 aborted/全挂 502/三家归一/过滤,全量 345 passed
This commit is contained in:
tzt
2026-09-05 09:14:31 +08:00
parent a9f3405970
commit f24f016e95
4 changed files with 345 additions and 13 deletions
+8
View File
@@ -21,11 +21,13 @@ _POOL_PATH = Path(__file__).resolve().parent.parent / "config" / "model_pool.jso
TIERS = ("local", "budget", "premium")
BACKENDS = ("mock", "llama_server", "openai")
ROLES = ("architect", "worker", "agent")
PROVIDERS = ("deepseek", "openai", "anthropic") # 代理层 usage 归一化用(D-P3)
# 池条目允许的字段(其余字段拒绝写入)
ENTRY_FIELDS = {
"id", "name", "tier", "backend", "base_url", "model", "api_key",
"price_in", "price_out", "temperature", "max_tokens", "enabled",
"provider", "in_hit_price", # 代理层扩展(D-P3):usage 归一化 / 按命中价选上游
}
# 单价默认值($/1M tokens);local 档为 0
@@ -177,6 +179,8 @@ class PoolStore:
backend = entry.get("backend", "openai")
if backend not in BACKENDS:
raise PoolError(f"backend 必须是 {BACKENDS} 之一")
if entry.get("provider") and entry["provider"] not in PROVIDERS:
raise PoolError(f"provider 必须是 {PROVIDERS} 之一")
tier = entry.get("tier", "budget")
if tier not in TIERS:
raise PoolError(f"tier 必须是 {TIERS} 之一")
@@ -212,6 +216,10 @@ class PoolStore:
"temperature": temperature,
"max_tokens": max_tokens,
"enabled": bool(entry.get("enabled", True)),
# 代理层扩展(D-P3):provider 缺省 openaiin_hit_price 缺省 = price_in × 1/30
"provider": (str(entry["provider"]) if entry.get("provider") else "openai"),
"in_hit_price": (float(entry["in_hit_price"]) if entry.get("in_hit_price") is not None
else round(price_in / 30.0, 6)),
}
@staticmethod