用于灾难恢复的 MSSQL 备份策略

MSSQL backup strategy for Disaster Recovery

我们的数据库是 MSSQL,我们目前在多节点集群中使用具有故障转移功能的高可用性组,因此冗余和备份的想法已经存在。

我们当前所有的服务器都在同一个位置;想象一个场景,地震摧毁了整个托管设施,然后我们就坐视不理了。

我正在探索一种灾难恢复 (DR) 策略,以便在不同的位置进行另一个 DR 备份,这样当发生这种情况时,我可以使用 DR 的备份集恢复整个数据库,同时最大限度地减少停机时间和数据需求如果可能的话,保证跟上会议纪要。

我已经阅读了 Microsoft 文档,但我没有真正看到详细讨论此问题的文档。

我需要一个最新的真正备份,我是否需要做这个完整备份(每天一次)和事务日志备份(每分钟一次),然后将其保存到其他不同的地理位置?您能否为我指出有关如何实现此目标的指南或最佳实践文档?

I'm exploring a disaster recover (DR) strategy to have another DR backup at a different location so when this happens I can bring back the entire database using the DR's backup set with a minimum down time and the data needs to be guarantee to be up to the minutes if possible

以下是您的选择,因为您已经有了可用性组。

  1. Multi sub-net WSFC You need to add additional node into WSFC, that work as DR Replica in availability group, consider it's another copy of secondary that you already have in same location but this copy laying different geographical location, since it's multi sub-net WSFC required attention for care-full Quorum configuration
  2. Log-Shipping 与多子网 WSFC 相比,这是一个简单的解决方案,并且易于管理。它基本上从当前可用性组中的主副本按计划进行日志备份,然后恢复到辅助副本。您可以拥有多个辅助副本,并且每个副本都位于不同的地理位置,具体取决于网络带宽。

I need a true backup that is up to the minutes, do I need to do this full backup (once every day) along with transaction log backup (one every minute) and then save it to the other different geological location? Can you point me to a guide or best practice documentation on how to achieve this?

This post at DBA.SE 会帮助你..