多子网始终开启可用性组（主节点和辅助节点都关闭，DR节点已启动）情况 [英] Multi Subnet Always On Availability group ( Both Primary and Secondary Nodes down, and DR Node is up) case
问题描述
亲爱的MSSQL专家，
Dear MSSQL Experts,
我们设置了多子网Always On设置主节点（同步）的2个节点和灾难恢复（异步）站点的一个节点，带有磁盘见证在PR网站。我需要知道如果PR站点的两个节点都关闭会发生什么？当
节点在10-30分钟之后出现时会发生什么，在这种情况下可以做些什么？如果计划内或计划外的活动/故障发生在PR网站上，应该采取什么预防措施？
We have setup Multi subnet Always On setup of 2 nodes at primary(synchronous) and one node at Disaster Recovery(Asynchronous) site with disk witness at PR site. I need to know what happens if both the nodes at PR site are down? What will happen when the nodes comes up after sometime say 10-30 minutes, what can be done in such situations? what precautions one should take if it is planned or unplanned activity/failure happens at PR site?
我发现我们可以强制故障转移到灾难恢复站点，但任何人都可以分享如何我可以使它与以前配置相同，即DR到PR。
I came across that we can force failover to DR site, but can anybody please share how can i make it same as it was previously configured i.e DR to PR .
我主要担心的是当两个节点都关闭时PR网站会发生什么，我们需要做什么它出现了同步。如果我弄错了，请纠正我。
My main concern is what happens to PR site when both nodes are down and do we need to do anything when it comes up for synchronization. Please correct me if i have mistaken.
谢谢，
Devendra
Devendra Yadav
Devendra Yadav
推荐答案
亲爱的MSSQL专家，
Dear MSSQL Experts,
我们已设置多子网始终在主节点（同步）上设置2个节点，在灾难恢复（异步）站点上设置一个节点，并在PR站点上显示磁盘见证。我需要知道如果PR站点的两个节点都关闭会发生什么？当
节点在10-30分钟之后出现时会发生什么，在这种情况下可以做些什么？如果计划内或计划外的活动/故障发生在PR网站上，应该采取什么预防措施？
We have setup Multi subnet Always On setup of 2 nodes at primary(synchronous) and one node at Disaster Recovery(Asynchronous) site with disk witness at PR site. I need to know what happens if both the nodes at PR site are down? What will happen when the nodes comes up after sometime say 10-30 minutes, what can be done in such situations? what precautions one should take if it is planned or unplanned activity/failure happens at PR site?
假设所有节点都投票，如果2个节点关闭在PR，你只剩下一个节点和磁盘仲裁，这不足以形成仲裁，WSFC将会崩溃。当节点在某个时间之后到来时，WSFC将上线，但是你会有停机时间。
Assuming all nodes are voting if 2 nodes are down at PR you are left with just one node and disk quorum which is not sufficient to form quorum and WSFC will come down. When nodes come after sometime the WSFC will come online then but you would have downtime.
在这种情况下，您可以强制仲裁在DR节点上联机，故障转移AG"允许数据丢失" ;并在DR节点上使数据库联机。
In such case you can force quorum to come online on DR node, failover AG with "allow data loss" and bring database online on DR node.
我发现我们可以强制故障转移到灾难恢复站点，但任何人都可以分享我怎样才能做到与先前配置相同，即DR到PR。
I came across that we can force failover to DR site, but can anybody please share how can i make it same as it was previously configured i.e DR to PR .
我主要担心的是，当两个节点都关闭时，PR网站会发生什么，当我们需要做什么时我们需要做什么同步。如果我弄错了，请纠正我。
My main concern is what happens to PR site when both nodes are down and do we need to do anything when it comes up for synchronization. Please correct me if i have mistaken.
参见 此博客使DR节点在线和数据库联机。请注意，您必须使用"强制故障转移允许数据丢失"来中断AG并使数据库联机。如果PR下降使WSFC失败，应用程序将会关闭。您需要使用上面的博客将WSFC在DR节点上联机。这样做
将打破AG并且PR上线后你将不得不重新配置AG
See this Blog to bring DR node online and database online. Please note you would have to break AG and bring database online by using "force failover allow data loss". If PR is down bringing WSFC down your case the application would be down. You need to use above blog to bring WSFC online on DR node. Doing this will break AG and after PR comes online you would have to reconfigure AG
这篇关于多子网始终开启可用性组（主节点和辅助节点都关闭，DR节点已启动）情况的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
