sql server 群集_SQL Server 2014 –安装群集实例

sql server 群集

Do you want to install a SQL Server 2014 clustered instance? Not sure how to proceed? What are the requisites? This series of articles will show this, step-by-step.

您是否要安装SQL Server 2014群集实例？不确定如何进行？有哪些条件？本系列文章将逐步展示这一点。

In this article I’m going to explain, step-by-step, how to install clustered SQL Server instance. The SQL Server version that we will be using is the 2014, or SQL Server 12, the latest SQL Server version, so far…

在本文中，我将逐步说明如何安装群集SQL Server实例。到目前为止，我们将使用SQL Server版本是2014，即SQL Server 12，即最新SQL Server版本。

In this environment, we have a cluster with two nodes, based on Windows Server 2012 R2 operating systems. This cluster is based in a domain called SQLCAFE.

在此环境中，我们有一个基于Windows Server 2012 R2操作系统的具有两个节点的群集。该群集基于一个称为SQLCAFE的域。

Regarding the cluster nodes, we have one network – 192.100.100.0/24 – connecting all the nodes and all other servers/computers. This is bases in my local lab, not a real production service. I opted to take advantage of the possibility to maintain a copy of the TempDB in a local disk, in order to have a batter performance, and added a SSD drive to each one of the nodes – local disk not shared. We also have more 4 shared storage, to support the data and log files, as well as the backups.

关于群集节点，我们有一个网络-192.100.100.0/24-连接所有节点和所有其他服务器/计算机。这是我本地实验室的基础，不是真正的生产服务。我选择利用在本地磁盘中维护TempDB副本的可能性，以提高性能，并为每个节点添加了一个SSD驱动器-本地磁盘不共享。我们还有4个共享存储，以支持数据和日志文件以及备份。

I tried to better simulate a real environment, during this installation I didn’t use a user without special permission in the domain, just to assure that we are following all the needed steps.

我试图更好地模拟真实环境，在此安装过程中，我没有使用没有域内特殊权限的用户，只是为了确保我们遵循所有必需的步骤。

In the picture bellow you can find a schema of what we will have in this step-by-step guide. Even if you have a cluster with more nodes, you can follow this guide, as the step to add the clustered instance to other nodes is the same, so it doesn’t matters if you are dealing with a threes nodes cluster or a 8 nodes one, the procedure is always the same.

在下面的图片中，您可以找到本逐步指南中所包含内容的架构。即使您的集群中有更多节点，也可以按照本指南进行操作，因为将集群实例添加到其他节点的步骤是相同的，因此，如果您要处理的是三节点集群或8节点，则没关系一，程序总是一样的。

Note: I recommend the entire reading of the article before start with the actual installation, this way you can find out something that you are missing, or some constraint that may suspend the installation.

注意：在开始实际安装之前，我建议您完整阅读本文，以这种方式，您可以发现所缺少的内容或某些可能会中止安装的约束。

The following prerequisites are needed in order to achieve a successfully clustered SQL Server instance installation:

为了成功完成群集SQL Server实例安装，需要满足以下先决条件：

SQL Server Engine service account: SQLCAFE\SVCSQLEngineSQL Server Engine服务帐户：SQLCAFE \ SVCSQLEngine
SQL Server Agent service account: SQLCAFE\SVCSQLAgentSQL Server代理服务帐户：SQLCAFE \ SVCSQLAgent
Computer Name: SQLSRVCLT01电脑名称：SQLSRVCLT01
Virtual IP (VIP): 192.100.100.50虚拟IP（VIP）： 192.100.100.50

Note: As a best practice, I create one service account for each SQL Server service, but this is not required. You can create one account to all the services if you want (SQL Server Engine, SQL Server Agent, SQL Server Integration Services, etc…).

注意：作为最佳实践，我为每个SQL Server服务创建一个服务帐户，但这不是必需的。如果需要，您可以为所有服务创建一个帐户（SQL Server引擎，SQL Server代理，SQL Server集成服务等）。

You can have more than one VIP, if you have different subnets. For example, if you have a dedicate network to perform backups (great!) you can assign an IP for the backups network and the front-end IP to the same computer name.

如果您具有不同的子网，则可以拥有多个VIP。例如，如果您有一个专用网络来执行备份（很棒！），则可以为备份网络分配一个IP，并将前端IP分配给同一计算机名。

The following should be assured for the SQL Server computer name:

对于SQL Server计算机名称，应确保以下几点：

The computer name “SQLSRVCLT01” was previously created in the domain and marked as disabled.先前在域中创建了计算机名称“ SQLSRVCLT01”，并将其标记为禁用。
The cluster computer – W2012R2CLT01 – should have full access to the SQL Server computer account – SQLSRVCLT01.

群集计算机W2012R2CLT01应该具有对SQL Server计算机帐户SQLSRVCLT01的完全访问权限。

In order to accomplish this, the following steps are needed in the domain level:

为此，在域级别需要执行以下步骤：
1. In the “Active Directory Users and Computers” tool, activate the “Advanced Features”, as shown:
  
  在“ Active Directory用户和计算机”工具中，激活“高级功能”，如下所示：
2. Now go to the Computers folder, right click the created SQL Server Computer and add the permission in the “Security” tab:
  
  现在转到“计算机”文件夹，右键单击创建SQL Server计算机，然后在“安全性”选项卡中添加权限：

谁在安装SQL Server？ ( Who is installing SQL Server? )

I see people using the SQL Server service account to install the instance, in this case I’m using the domain user SQLCAFE\murilo to install the SQL Server clustered instance. This user is a normal user in the domain – not domain admin and has no special grants. The user SQLCAFE\murilo is part of the Administrators group on in both of the cluster’s node: W2012R2SRV01 and W2012R2SRV02. I didn’t set any permissions to both SQL Server service accounts any nodes of the cluster. We will see what is going to happen in the end of the installation…

我看到有人使用SQL Server服务帐户安装实例，在这种情况下，我使用域用户SQLCAFE \ murilo安装SQL Server群集实例。该用户是域中的普通用户–不是域管理员，并且没有特殊的授予。用户SQLCAFE \ murilo是群集的以下两个节点中的Administrators组的一部分：W2012R2SRV01和W2012R2SRV02。我没有为两个SQL Server服务帐户设置任何权限，也没有为群集的任何节点设置权限。我们将看到安装结束时将会发生什么……

存储 ( Storage )

For a clustered instance we need to use shared storage. The disk for the TempDB files and Buffer Pool Extension disk (if used) are exceptions. On both cases we need to have the local disks attached to each node and mapped to the same drive letter/mountpoint path.

对于集群实例，我们需要使用共享存储。 TempDB文件的磁盘和缓冲池扩展磁盘（如果使用）是例外。在这两种情况下，我们都需要将本地磁盘连接到每个节点并映射到相同的驱动器号/挂载点路径。

For this instance, I’m creating the following disks:

对于此实例，我正在创建以下磁盘：

Assigned to d分配到d rive 里沃 letter 字母 E: Disk for SQL Server Binaries (local)E：盘以供SQL Server二进制文件（本地）
Assigned to drive letter F: Disk for Data Files (shared)
分配到D 里沃字母F：盘以供数据文件（共享）
Assigned to drive letter G: Disk for Log Files (Shared)
分配给d 里沃字母G：磁盘日志文件（共享）
Assigned to drive letter H: Disk for Backups (Shared)
分配给d 里沃字母H：磁盘备份（共享）
Assigned to d分配到d rive 里沃 letter I信我 : Disk for TempDB (local)：磁盘tempdb的（本地）

All the shared disks need to be shown as “Available Storage” in the Failover Cluster Manager, as you can see:

您将在故障转移群集管理器中将所有共享磁盘显示为“可用存储”：

I formatted the SQL Server disks using an Allocation Unit of 64K, as the performance best practice states.

根据性能最佳实践的说明，我使用64K的分配单元格式化了SQL Server磁盘。

安装过程 ( The Installation Process )

Now that we have all the prerequisites, we can start the installation. As said, I’m using my personal domain account to log into the cluster nodes and start the SQL Server installation.

现在我们已经具备所有先决条件，我们可以开始安装了。如上所述，我正在使用个人域帐户登录群集节点并开始SQL Server安装。

The approach to install the clustered instance is run the setup in the first node, and pick the option to install a new SQL Server clustered instance. After the initial installation, in the first node, we need to run the setup in all other nodes that we want to install the SQL Server clustered instance, and pick the option to add a node to an existing clustered SQL Server instance. We will walk through all of this in the continuation of this article.

安装群集实例的方法是在第一个节点中运行安装程序，然后选择安装新SQL Server群集实例的选项。初始安装后，在第一个节点中，我们需要在要安装SQL Server群集实例的所有其他节点中运行安装程序，并选择将节点添加到现有群集SQL Server实例的选项。我们将在本文的后续部分中逐步介绍所有这些内容。

从第一个节点开始 ( Starting from the first node )

I choose the node W2012R2SRV01 to start installing the instance. The reason for that? There’s no reason, only a matter of choice