probe v1.0.2 container submit test

main.py

@@ -1,15 +1,14 @@
-"""ModelHub XC 适配智能体 · 只读探针骨架 (probe-only)
+"""ModelHub XC 适配智能体 · 只读探针 + 容器内提交测试 v1.0.2
 
-用途:
-  1) 验证平台「建仓 -> Kaniko 构建 -> 部署 -> /health -> 运行/停止」整条链路。
-  2) 零副作用观测平台注入的 STRATEGY_ID / EXTERNAL_SERVICE_TOKEN 是否为有效凭证
-     (启动时用注入 token 调一次只读 GET /api/adapt/task/page,看返回 code:
-      0 = 凭证有效; 40100 = 未登录/凭证无效)。
+关键实验:在容器里用注入的 EXTERNAL_SERVICE_TOKEN(=tmp) + STRATEGY_ID 真调一次 task/add,
+验证平台是否靠「请求来自容器 pod」放行智能体提交(浏览器/脚本提交会被 60014 挡)。
 
-严格约束:
-  - 只服务 GET /health 与 GET /,绝不提交任何验证任务 (不调 task/add, 不调 build-config)。
-  - token 只从环境读,绝不打印其值/落盘;探针只调只读 GET;只记录 http/code/计数,不打印返回体内容(防 PII)。
-  - 纯 stdlib 零依赖;正确处理 SIGTERM 优雅停机。
+诊断输出:
+  - dump 所有 env 的名字+长度(不打值)。
+  - EXTERNAL_SERVICE_TOKEN 值(短则原样、长则打码)。
+  - 容器内 build-config + task/add(真实模型) 的返回 code/msg。
+
+约束:token 值不打(短占位除外);只提交这一次测试(真实模型可能进队列);SIGTERM 优雅停机。
 """
 import json
 import os
@@ -21,62 +20,77 @@ import urllib.request
 from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
 
 MAIN = os.getenv("MAIN_HOST", "https://modelhub.org.cn")
-TOKEN = os.getenv("EXTERNAL_SERVICE_TOKEN", "")     # 平台注入的服务 token
-STRATEGY_ID = os.getenv("STRATEGY_ID", "")          # 平台注入的自身策略 id
+TOKEN = os.getenv("EXTERNAL_SERVICE_TOKEN", "")
+STRATEGY_ID = os.getenv("STRATEGY_ID", "")
 PORT = int(os.getenv("PORT", "8080"))
 
+TEST_MODEL = os.getenv("TEST_MODEL", "https://www.modelscope.cn/models/mradermacher/TrialSpace-1225-GGUF")
+GPU, FW, TT = os.getenv("GPU", "MetaX_c-500"), os.getenv("FW", "vllm"), os.getenv("TT", "text-generation")
+
+HARDCODED_CFG = (
+    "framework: vllm\napi: completion\nlang: en\nmax_model_len: 4096\nmax_tokens: 1024\n"
+    "temperature: 0.7\nrepetition_penalty: 1.1\ntop_p: 0.9\nsut_config:\n  gpu_num: 1\n  values:\n"
+    "    command:\n    - /opt/conda/bin/vllm\n    - serve\n    - /model\n    - --port\n    - '20644'\n"
+    "    - --served-model-name\n    - llm\n    - --max-model-len\n    - '4096'\n"
+    "    - --gpu-memory-utilization\n    - '0.9'\n    - -tp\n    - '1'\n    - --enforce-eager\n"
+    "    - --trust-remote-code\nref_config:\n  gpu_num: 1\n  values:\n    command:\n    - vllm\n"
+    "    - serve\n    - /model\n    - --port\n    - '80'\n    - --served-model-name\n    - llm\n"
+    "    - --max-model-len\n    - '4096'\n    - -tp\n    - '1'\n    - --enforce-eager\n    - --trust-remote-code\n"
+)
+
 shutdown = threading.Event()
-PROBE = {"done": False, "results": []}              # 探针结果,暴露在 GET / (不含任何 token 值)
+PROBE = {"done": False, "results": []}
 
 
 def log(msg: str) -> None:
     print(f"[{time.strftime('%Y-%m-%d %H:%M:%S')}] {msg}", flush=True)
 
 
-def _readonly_get(path: str, headers: dict, timeout: int = 20) -> dict:
-    """只读 GET,返回 {http, code, count};绝不返回/记录响应体内容。"""
-    req = urllib.request.Request(MAIN + path, method="GET", headers=headers)
+def _mask(v: str) -> str:
+    return v if len(v) <= 8 else (v[:2] + "…" + v[-2:])
+
+
+def _req(method: str, path: str, headers: dict, body=None, timeout: int = 30) -> dict:
+    data = body.encode() if isinstance(body, str) else body
+    req = urllib.request.Request(MAIN + path, data=data, method=method, headers=headers)
     try:
         with urllib.request.urlopen(req, timeout=timeout) as r:
-            body = r.read().decode("utf-8", "ignore")
+            txt = r.read().decode("utf-8", "ignore")
             status = r.status
     except urllib.error.HTTPError as e:
-        body = e.read().decode("utf-8", "ignore")
+        txt = e.read().decode("utf-8", "ignore")
         status = e.code
     except Exception as e:  # noqa: BLE001
-        return {"http": -1, "err": str(e)[:120]}
-    code = count = None
+        return {"http": -1, "err": str(e)[:100]}
     try:
-        j = json.loads(body)
-        code = j.get("code")
-        d = j.get("data")
-        if isinstance(d, dict):
-            count = d.get("total")
-            if count is None and isinstance(d.get("records"), list):
-                count = len(d["records"])
-        elif isinstance(d, list):
-            count = len(d)
+        j = json.loads(txt)
+        return {"http": status, "code": j.get("code"), "msg": (j.get("message") or "")[:100],
+                "data": j.get("data")}
     except Exception:  # noqa: BLE001
-        pass
-    return {"http": status, "code": code, "count": count}
+        return {"http": status, "raw": txt[:100]}
 
 
 def probe() -> None:
-    """一次性只读探针:验证注入凭证有效性 + 记录 STRATEGY_ID。"""
-    log(f"probe start | strategy_id={STRATEGY_ID or 'MISSING'} | "
-        f"token_present={bool(TOKEN)} | token_len={len(TOKEN)}")
-    path = "/api/adapt/task/page?current=1&pageSize=1"
-    trials = [
-        ("Xc-Token", {"Xc-Token": TOKEN}),
-        ("Authorization-Bearer", {"Authorization": "Bearer " + TOKEN}),
-        ("Authorization-raw", {"Authorization": TOKEN}),
-    ]
-    for name, headers in trials:
-        res = {"skip": "no-token"} if not TOKEN else _readonly_get(path, headers)
-        PROBE["results"].append({"auth": name, **res})
-        log(f"probe {name} -> {json.dumps(res, ensure_ascii=False)}")
+    env_lens = {k: len(str(os.environ.get(k, ""))) for k in sorted(os.environ)}
+    log("env keys+lens: " + json.dumps(env_lens, ensure_ascii=False))
+    log(f"strategy_id={STRATEGY_ID or 'MISSING'} | EXTERNAL_SERVICE_TOKEN value={_mask(TOKEN)!r} len={len(TOKEN)}")
+
+    # 只读:token 有效性
+    page = _req("GET", "/api/adapt/task/page?current=1&pageSize=1", {"Xc-Token": TOKEN})
+    log("EST-token task/page -> " + json.dumps({k: page.get(k) for k in ("http", "code")}, ensure_ascii=False))
+
+    # 关键:容器内 build-config + task/add(真实模型)
+    bc = _req("POST", f"/api/adapt/task/build-config?gpuType={GPU}&framework={FW}&taskType={TT}", {"Xc-Token": TOKEN})
+    log("build-config(EST-token) -> " + json.dumps({k: bc.get(k) for k in ("http", "code", "msg")}, ensure_ascii=False))
+    cfg = bc.get("data") if isinstance(bc.get("data"), str) and bc.get("data") else HARDCODED_CFG
+    body = json.dumps({"modelAddress": TEST_MODEL, "taskType": TT, "targetGpu": GPU,
+                       "framework": FW, "strategyId": STRATEGY_ID, "configParams": cfg})
+    ta = _req("POST", "/api/adapt/task/add", {"Xc-Token": TOKEN, "Content-Type": "application/json"}, body)
+    log("SUBMIT-TEST task/add(EST-token+strategyId, real model) -> " + json.dumps(ta, ensure_ascii=False))
+    PROBE["results"] = [{"task_page": page.get("code")}, {"build_config": bc.get("code")},
+                        {"task_add": ta.get("code"), "msg": ta.get("msg")}]
     PROBE["done"] = True
-    log("probe done | 判读: code=0 => 该头凭证有效; code=40100 => 未登录/凭证无效")
+    log("probe done | task/add code=0 => 容器提交通; 60014 => 容器来源也被判脚本; 401 => token 无效")
 
 
 class Handler(BaseHTTPRequestHandler):
@@ -93,13 +107,7 @@ class Handler(BaseHTTPRequestHandler):
             self._json({"status": "ok"})
             return
         if self.path == "/":
-            self._json({
-                "name": "huni-probe-agent",
-                "mode": "probe-only",
-                "strategy_id_present": bool(STRATEGY_ID),
-                "token_present": bool(TOKEN),
-                "probe": PROBE,
-            })
+            self._json({"name": "huni-probe-agent", "mode": "submit-test-v1.0.2", "probe": PROBE})
             return
         self._json({"error": "not found"}, 404)
 
@@ -119,7 +127,7 @@ def main() -> None:
 
     server = ThreadingHTTPServer(("0.0.0.0", PORT), Handler)
     server.timeout = 1
-    log(f"probe-only agent listening on 0.0.0.0:{PORT}")
+    log(f"probe agent v1.0.2 listening on 0.0.0.0:{PORT}")
     while not shutdown.is_set():
         server.handle_request()
     server.server_close()