Mongoose间歇性无法连接MongoDB Atlas集群问题求助

我在Vercel上运行Express服务器，使用Mongoose 7.6.13版本，遇到间歇性数据库连接失败问题，该问题通常在函数闲置一段时间后出现。由于服务大部分时间能正常运行，已排除防火墙/IP白名单问题。使用的是Mongo Atlas M10套餐，服务器负载极低，请求协助排查。

连接代码 #

const express = require("express");
const mongoose = require("mongoose");

mongoose.connect("mongodb+srv://prod-database-vercel:ABC@production.ABC.mongodb.net/ABC",{
  useNewUrlParser: true,
  useUnifiedTopology: true,
}).catch(e => {
  throw new Error('error caught from mongoose connect', { cause: e })
});

间歇性错误日志 #

Unhandled Rejection: MongooseServerSelectionError: Could not connect to any servers in your MongoDB Atlas cluster. One common reason is that you're trying to access the database from an IP that isn't whitelisted. Make sure your current IP address is on your Atlas cluster's IP whitelist: https://www.mongodb.com/docs/atlas/security-whitelist/
at _handleConnectionErrors (/var/task/backend/node_modules/mongoose/lib/connection.js:809:11)
at NativeConnection.openUri (/var/task/backend/node_modules/mongoose/lib/connection.js:784:11)
at processTicksAndRejections (node:internal/process/task_queues:95:5) {
reason: TopologyDescription {
type: 'ReplicaSetNoPrimary',
servers: Map(3) {
'ABC-00-00.ABC.mongodb.net:27017' => [ServerDescription],
'ABC-shard-00-01.ABC.mongodb.net:27017' => [ServerDescription],
'ABC-shard-00-02.ABC.mongodb.net:27017' => [ServerDescription]
},
stale: false,
compatible: true,
heartbeatFrequencyMS: 10000,
localThresholdMS: 15,
setName: 'atlas-wnvhrz-shard-0',
maxElectionId: null,
maxSetVersion: null,
commonWireVersion: 0,
logicalSessionTimeoutMinutes: null
},
code: undefined
}
Shutting down the server for Could not connect to any servers in your MongoDB Atlas cluster. One common reason is that you're trying to access the database from an IP that isn't whitelisted. Make sure your current IP address is on your Atlas cluster's IP whitelist: https://www.mongodb.com/docs/atlas/security-whitelist/
shutting down the server for unhandle promise rejection
Node.js process exited with exit status: 1. The logs above can help with debugging the issue.

排查方案 #

• 优化Mongoose连接配置
  闲置后连接失效通常是因为Atlas的空闲连接超时或Vercel函数的连接回收。调整连接参数适配Serverless环境：

  mongoose.connect("mongodb+srv://prod-database-vercel:ABC@production.ABC.mongodb.net/ABC",{
    socketTimeoutMS: 45000, // 延长套接字超时时间
    connectTimeoutMS: 30000, // 缩短连接超时，快速失败并重连
    heartbeatFrequencyMS: 5000, // 缩短心跳间隔，及时检测连接状态
    maxPoolSize: 10, // 限制连接池大小，适配Serverless资源
    // Mongoose 7.x已默认启用useNewUrlParser和useUnifiedTopology，可移除这两个参数
  });
  

• 添加重连逻辑，避免进程崩溃
  当前代码在连接失败时直接抛出错误导致进程退出，改为监听连接事件实现自动重连：

  const db = mongoose.connection;
  
  db.on('error', (err) => {
    console.error('MongoDB连接错误:', err);
  });
  
  db.on('disconnected', () => {
    console.log('MongoDB连接断开，尝试重连...');
    mongoose.connect("mongodb+srv://prod-database-vercel:ABC@production.ABC.mongodb.net/ABC", {
      socketTimeoutMS: 45000,
      connectTimeoutMS: 30000,
      heartbeatFrequencyMS: 5000,
      maxPoolSize: 10
    });
  });
  
  db.once('open', () => {
    console.log('MongoDB连接成功');
  });
  

• 适配Vercel Serverless特性
  Vercel函数闲置后会休眠，冷启动时原有连接可能已失效。建议：

  • 在请求处理前检查连接状态，必要时触发重连
  • 使用Vercel官方的MongoDB集成，优化连接管理流程

• 查看Atlas连接日志
  登录MongoDB Atlas控制台，查看集群的连接日志，确认闲置后连接断开的具体原因（如Atlas主动关闭、网络波动等），进一步定位问题。

内容的提问来源于stack exchange，提问作者Mohammad Ali