MongoDB免费集群突然无法连接，部署变更异常求助

问题现象 #

• MongoDB Compass连接时报错：

2281984:error:10000438:SSL routines:OPENSSL_internal:TLSV1_ALERT_INTERNAL_ERROR:..\..\third_party\boringssl\src\ssl\tls_record.cc:592:SSL alert number 80

• Atlas控制台顶部显示：We are deploying your changes (current action: creating a plan)
• Node.js（Mongoose）连接同样失败，报错信息如下：

MongoDB connection error: 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 (E:\1. Programiranje\Web projekti\Balkanflix\Balkanflix\server\node_modules\mongoose\lib\connection.js:788:11)
    at NativeConnection.openUri (E:\1. Programiranje\Web projekti\Balkanflix\Balkanflix\server\node_modules\mongoose\lib\connection.js:763:11)
    at async connectDB (E:\1. Programiranje\Web projekti\Balkanflix\Balkanflix\server\config\db.js:5:9) {
  reason: TopologyDescription {
    type: 'ReplicaSetNoPrimary',
    servers: Map(3) {
      'ac-fiu19tn-shard-00-01.bmscihv.mongodb.net:27017' => [ServerDescription],
      'ac-fiu19tn-shard-00-02.bmscihv.mongodb.net:27017' => [ServerDescription],
      'ac-fiu19tn-shard-00-00.bmscihv.mongodb.net:27017' => [ServerDescription]
    },
    stale: false,
    compatible: true,
    heartbeatFrequencyMS: 10000,
    localThresholdMS: 15,
    setName: 'atlas-jj3mwx-shard-0',
    maxElectionId: null,
    maxSetVersion: null,
    commonWireVersion: 0,
    logicalSessionTimeoutMinutes: null
  },
  code: undefined
}

• 同事尝试连接也失败，排除本地机器问题；已在MongoDB社区搜索无相关解决方案，特此求助。

可行的排查与解决方法 #

• 等待部署完成：Atlas显示正在部署变更时，集群处于不稳定状态，免费集群部署通常需要10-30分钟，耐心等待流程结束后再尝试连接。
• 检查节点状态：进入Atlas集群的「Overview」页面，确认所有分片节点是否正常运行，有无宕机或维护标记。
• 重新生成连接字符串：部署完成后仍无法连接的话，在控制台重新生成Compass和Mongoose的连接字符串，替换现有配置后重试。
• 验证IP白名单：虽然多人连接失败大概率不是单个IP问题，但仍需确认「Network Access」中的白名单是否包含需要访问的IP段，可临时添加0.0.0.0/0（测试后及时移除，避免安全风险）验证。
• 提交官方工单：以上方法无效时，通过Atlas控制台的「Support」入口提交工单，提供报错信息和集群状态截图，请求官方排查。

内容的提问来源于stack exchange，提问作者Strahinja