多Worker守护 + 告警服务器名 + 清理PHP残留

- main.py: Redis主Worker选举，防止多Worker重复执行后台任务
- agent.py: 告警/恢复广播使用真实服务器名(查MySQL)替代"server-{id}"
- sync_v2.py: 修复browse目录解析死代码，正确取parts[8]
- install.html: 移除过时install.php引用，更新为"删除.env重装"
- nginx配置: 移除PHP-FPM路由(install.html纯静态无需PHP)
- schedule_runner: 清理未使用的get_redis_sync导入

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

server/api/agent.py

@@ -115,6 +115,15 @@ async def receive_heartbeat(
         logger.error(f"Redis heartbeat write failed for server {server_id}: {e}")
         # Redis write failed — still continue with MySQL update
 
+    # ── 1b. Look up server name for alert broadcasts ──
+    server_name = f"server-{server_id}"
+    try:
+        server = await service.get_server(server_id)
+        if server:
+            server_name = server.name
+    except Exception:
+        pass
+
     # ── 2. Alert detection ──
     alert_thresholds = {
         "cpu": settings.CPU_ALERT_THRESHOLD,
@@ -126,7 +135,7 @@ async def receive_heartbeat(
         alerts = _detect_alerts(system_info, alert_thresholds)
         for alert_type, alert_value in alerts:
             logger.warning(f"Alert: server {server_id} {alert_type}={alert_value}%")
-            await broadcast_alert(server_id, alert_type, alert_value, f"server-{server_id}")
+            await broadcast_alert(server_id, alert_type, alert_value, server_name)
 
             # Track active alert in Redis
             try:
@@ -145,7 +154,7 @@ async def receive_heartbeat(
                 recoveries = _detect_recovery(system_info, prev_alerts, alert_thresholds)
                 for metric, value in recoveries:
                     logger.info(f"Recovery: server {server_id} {metric}={value}%")
-                    await broadcast_recovery(server_id, metric, value, f"server-{server_id}")
+                    await broadcast_recovery(server_id, metric, value, server_name)
 
                     # Remove recovered alert from Redis tracking
                     for prev in prev_alerts: