Windows 2012 R2 / SQL 2016 /块存储 - 可用性选项 [英] Windows 2012 R2/SQL 2016/Block Storage - Availability Options

问题描述

您好，

我的任务是构建一个高度可用的SQL服务器。此服务器将是用于Citrix DaaS环境的数据库。始终在SQL上不一定是必需的，但是延长的中断是不可接受的。

我有2个节点：节点1和节点2 - 我在两个节点上都配置了MS Fail over cluster。 / p>

我已经向两个节点提供了500GB的块存储。

我已经配置了我的集群并经过了验证，但验证说明： ;没有找到要执行集群验证测试的磁盘"

磁盘在群集管理控制台中处于联机状态。

我的问题，我是如果我使用块存储，需要有一个故障转移群集？我的驱动器是一个块存储卷，它有自己的生命周期管理。存储在块存储中的数据库文件可以在
灾难的情况下轻松移动到另一个SQL Server。

使用块存储的恢复能力是否有任何优势MS群集的可用性？ 

非常感谢，

解决方案

¥b $ b驱动器数量这些节点上是本地磁盘。这些不是共享存储。这就是您遇到错误的原因 "没有找到执行集群验证测试的磁盘"在集群验证报告中。

SQL FCI的工作方式是，所有数据库都存储在共享存储中，当从Node1到Node2发生故障转移时，磁盘也从Node1移动到Node2和SQL服务器从这些磁盘中选择所有内容。

SQL二进制文件安装在本地磁盘上。

SQL FCI必须要求共享存储。如果您使用块存储而不使用SQL FCI，则故障转移过程可能会更长，但是如果运行SQL FCI，它将只有几秒钟并且对用户透明。

谢谢，

Hello,

I have been tasked with building a Highly Available SQL server. This server will be the database used for a Citrix DaaS environment. Always on SQL is not necessarily required, but an extended outage is not acceptable.

I have 2 nodes: Node 1 and Node 2 - I have configured MS Fail over cluster on both Nodes.

I have available 500GB Block storage that has been presented to both nodes.

I have configured my cluster and gone through validation, but validation states: "No disks were found on which to perform cluster validation tests"

The disks are online in the Cluster management console.

My question, do I need to have a failover cluster if I am using block storage? My drive is a block storage volume that has its own life cycle management. Database files that are stored in block storage can easily be moved to another SQL Server in case of disaster.

Is there any advantage in using the recover ability of block storage and the availability of MS clustering? 

Many Thanks,

解决方案

Hi,
The drives\Disks you have on those Nodes are local disks. These are not shared storage. That is why you got error "No disks were found on which to perform cluster validation tests" during cluster validation report.

The way SQL FCI work is, all databases are stored in shared storage and when failover happen from Node1 to Node2, the disks are also moved from Node1 to Node2 and SQL server picks everything from those disks.

SQL binaries are installed on Local disks.

It is MUST requirement for SQL FCI to have shared storage. If you using block storage and not going for SQL FCI, the failover process might be longer however in case of a running SQL FCI, it will only few seconds and transparent to users.

Thanks,

这篇关于Windows 2012 R2 / SQL 2016 /块存储 - 可用性选项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！