feat: 综合平台services多服务支持+SW网络优先缓存修复+适配层/定时任务完善

agent/collector.py

@@ -34,15 +34,21 @@ def collect_metrics() -> dict:
 
 
 def _get_ip() -> str:
-    """获取本机出口 IP（不实际发包）"""
+    """获取本机出口 IP（UDP 不实际发包；设超时避免无网环境阻塞）"""
+    s = None
     try:
         s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
+        s.settimeout(2)  # 无网络环境下 connect 可能阻塞，加超时保护
         s.connect(("8.8.8.8", 80))
-        ip = s.getsockname()[0]
-        s.close()
-        return ip
+        return s.getsockname()[0]
     except Exception:
         return ""
+    finally:
+        if s is not None:
+            try:
+                s.close()
+            except Exception:
+                pass
 
 
 def collect_server_info() -> dict:

agent/reporter.py

@@ -4,10 +4,12 @@
   VPS_MANAGER_URL  中心地址，如 http://100.89.0.11:8000
   VPS_ASSET_ID     本机对应的资产 ID
   VPS_AGENT_KEY    Agent Key（中心配置了 AGENT_KEY 时必填）
+  VPS_AGENT_RETRY  上报失败重试次数（默认 2，指数退避）
 """
 
 import os
 import sys
+import time
 
 import httpx
 
@@ -18,6 +20,7 @@ def main() -> int:
     url = os.environ.get("VPS_MANAGER_URL", "").rstrip("/")
     asset_id = os.environ.get("VPS_ASSET_ID", "")
     key = os.environ.get("VPS_AGENT_KEY", "")
+    max_retry = int(os.environ.get("VPS_AGENT_RETRY", "2"))
 
     if not url or not asset_id:
         print("错误：需设置 VPS_MANAGER_URL 和 VPS_ASSET_ID")
@@ -31,16 +34,23 @@ def main() -> int:
     }
     headers = {"X-Agent-Key": key} if key else {}
 
-    try:
-        resp = httpx.post(
-            f"{url}/api/agent/report", json=payload, headers=headers, timeout=30
-        )
-        print(f"上报完成 HTTP {resp.status_code}: {resp.text}")
-        resp.raise_for_status()
-        return 0
-    except Exception as e:  # noqa: BLE001
-        print(f"上报失败: {e}")
-        return 1
+    for attempt in range(max_retry + 1):
+        try:
+            resp = httpx.post(
+                f"{url}/api/agent/report", json=payload, headers=headers, timeout=30
+            )
+            print(f"上报完成 HTTP {resp.status_code}: {resp.text}")
+            resp.raise_for_status()
+            return 0
+        except Exception as e:  # noqa: BLE001
+            if attempt < max_retry:
+                wait = 2 ** attempt
+                print(f"上报失败（{attempt + 1}/{max_retry + 1}），{wait}s 后重试: {e}")
+                time.sleep(wait)
+            else:
+                print(f"上报失败（已重试 {max_retry} 次）: {e}")
+                return 1
+    return 1
 
 
 if __name__ == "__main__":