fix: image_daemon 加心跳日志+异常捕获+进程守护自动重启

entrypoint.sh

@@ -268,12 +268,22 @@ echo "   Gateway PID: $GATEWAY_PID"
 sleep 5
 
 # ============================================
-# 启动图片预处理守护进程（后台）
+# 启动图片预处理守护进程（后台，带自动重启守护）
 # ============================================
-echo "🖼️ 启动图片预处理守护进程..."
-python3 /app/image_daemon.py &
+echo "🖼️ 启动图片预处理守护进程（带自动重启）..."
+(
+  RESTART_COUNT=0
+  while true; do
+    RESTART_COUNT=$((RESTART_COUNT + 1))
+    echo "[image_daemon_guard] 🚀 启动 image_daemon (第 ${RESTART_COUNT} 次)"
+    python3 /app/image_daemon.py
+    EXIT_CODE=$?
+    echo "[image_daemon_guard] ⚠️ image_daemon 退出 (code=${EXIT_CODE}), 3 秒后重启..."
+    sleep 3
+  done
+) &
 IMAGE_DAEMON_PID=$!
-echo "   Image Daemon PID: $IMAGE_DAEMON_PID"
+echo "   Image Daemon Guard PID: $IMAGE_DAEMON_PID"
 
 # ============================================
 # 启动状态监控网页（前台，端口 7860）
@@ -282,3 +292,4 @@ echo "📊 启动状态监控网页 (端口 7860)..."
 exec python3 /app/status_page.py
 
 
+

image_daemon.py

@@ -41,38 +41,46 @@ def get_tenant_token():
 
 def get_bot_chats(token):
     """获取 bot 参与的所有聊天"""
-    headers = {"Authorization": f"Bearer {token}"}
-    resp = requests.get(f"{FEISHU_BASE}/im/v1/chats",
-        headers=headers, params={"page_size": 20}, timeout=10)
-    data = resp.json()
-    code = data.get("code", -1)
-    if code != 0:
-        msg = data.get("msg", "unknown")
-        log(f"❌ 获取聊天列表失败 (code={code}): {msg}")
-        log(f"   需要权限: im:chat / im:chat:readonly")
+    try:
+        headers = {"Authorization": f"Bearer {token}"}
+        resp = requests.get(f"{FEISHU_BASE}/im/v1/chats",
+            headers=headers, params={"page_size": 20}, timeout=10)
+        data = resp.json()
+        code = data.get("code", -1)
+        if code != 0:
+            msg = data.get("msg", "unknown")
+            log(f"❌ 获取聊天列表失败 (code={code}): {msg}")
+            log(f"   需要权限: im:chat / im:chat:readonly")
+            return []
+        items = data.get("data", {}).get("items", [])
+        return [c["chat_id"] for c in items]
+    except Exception as e:
+        log(f"❌ get_bot_chats 异常: {type(e).__name__}: {e}")
         return []
-    items = data.get("data", {}).get("items", [])
-    return [c["chat_id"] for c in items]
 
 def get_recent_messages(token, chat_id, limit=10):
     """获取聊天中最近的消息"""
-    headers = {"Authorization": f"Bearer {token}"}
-    resp = requests.get(f"{FEISHU_BASE}/im/v1/messages",
-        headers=headers,
-        params={
-            "container_id_type": "chat",
-            "container_id": chat_id,
-            "sort_type": "ByCreateTimeDesc",
-            "page_size": limit
-        }, timeout=10)
-    data = resp.json()
-    code = data.get("code", -1)
-    if code != 0:
-        msg = data.get("msg", "unknown")
-        log(f"❌ 获取消息失败 chat={chat_id} (code={code}): {msg}")
-        log(f"   需要权限: im:message / im:message:readonly")
+    try:
+        headers = {"Authorization": f"Bearer {token}"}
+        resp = requests.get(f"{FEISHU_BASE}/im/v1/messages",
+            headers=headers,
+            params={
+                "container_id_type": "chat",
+                "container_id": chat_id,
+                "sort_type": "ByCreateTimeDesc",
+                "page_size": limit
+            }, timeout=10)
+        data = resp.json()
+        code = data.get("code", -1)
+        if code != 0:
+            msg = data.get("msg", "unknown")
+            log(f"❌ 获取消息失败 chat={chat_id} (code={code}): {msg}")
+            log(f"   需要权限: im:message / im:message:readonly")
+            return []
+        return data.get("data", {}).get("items", [])
+    except Exception as e:
+        log(f"❌ get_recent_messages 异常: {type(e).__name__}: {e}")
         return []
-    return data.get("data", {}).get("items", [])
 
 def extract_image_keys(msg):
     """从消息中提取所有 image_key"""
@@ -225,9 +233,11 @@ def main():
 
             chats = get_bot_chats(token)
 
-            # 首次和每 100 轮打印状态
-            if cycle == 1 or cycle % 100 == 0:
-                log(f"📊 轮询中... 找到 {len(chats)} 个聊天 (第 {cycle} 轮)")
+            # 每轮打印心跳（每 20 轮详细，其余简短）
+            if cycle == 1 or cycle % 20 == 0:
+                log(f"💓 轮询中... {len(chats)} 个聊天, 已处理 {len(processed)} 条 (第 {cycle} 轮)")
+            else:
+                log(f"💓 #{cycle} ok, {len(chats)} chats")
 
             if not chats and cycle == 1:
                 log("⚠️ 未找到任何聊天！可能原因:")