A Brief Introduction to SQL Server Failover Clustering

Stacy Brown ~ Modified: September 3rd, 2015 ~ High Availability ~ 5 Minutes Reading

Introduction

SQL Server failover clustering is the practice of maintaining a cluster of SQL Servers, in order to have high data availability. This helps the organization to access data from other servers at downtime of any one server. Each server in the cluster is known as a node. Let’s see the installation procedure of SQL Server Failover Cluster instance.

How to Install a Failover Cluster?

The failover cluster can be installed using the SQL server setup wizard. To install the setup you need a domain account with administrator rights. The right must include permission to log on as a service and should act as a part of OS on all nodes on the cluster. Some of the requirements that should be met for installation of a SQL server failover cluster instance are as follows.

For the installation and configuration of the SQL server failover clustering uses SQL Server setup.
Identify cluster disk resource, IP addresses, network name, etc. needed to create a failover cluster instance.
Use separate Windows Server Failover Cluster group for each cluster instance that you need to configure. The configuration must be carried out before the SQL Server Setup program uses the Windows Cluster Administrator to carry out the same.
The system should meet the minimum requirements for the cluster setup.
Addition or removal of nodes from a failover cluster can be carried out without affecting other nodes.
All the nodes in a cluster must be of the same platform, and must run the same operating system, either 32-bit or 64-bit of same version.
One can specify multiple IP addresses for each subnet of the clusters. AND dependency is specified if the IP addresses fall on the same subnet. Else if the IP addresses fall in different subnet the dependency is set to OR.

SQL Server Failover Clustering Installing Methods

Method 1: Enterprise/Advanced Installation Procedure

The two steps involved in this procedure are

1. Run SQL server setup with its preparing failover cluster functionality on each node that is the part of the cluster. This is done in order to prepare the nodes ready for clustering.

2. Run the setup on these prepared nodes that own a shared disk with complete failover functionality. This configures a complete failover cluster setup instance. Here an operational SQL server Failover Cluster instance in created.

Method 2: Using Add Note

This integrated installation method with add note adds a single node to the cluster at each level. This procedure is as follows

1. Configure a single node Cluster Instance at the first step. This may be a fully functional SQL server failover cluster.

2. To add a new node, Run the setup on each node to be added to the failover cluster with add Node functionality.

NOTE:

All the nodes added to the failover cluster must maintain same OS drive letter for SQL Server installation locations. The add Node method can be used to additional nodes to a failover cluster that have been already created.

The IP address is set or changed during any of the following actions including installation, addition or removal a new node, etc. Both IPV4 and IPV6 are supported in the SQL Server failover clustering but, both are considered as different subnets.

Add a Node

In order To add a node to an existing SQL Server cluster, one must run Server Setup on the node that is to be added to the instance. Make sure that setup is not run on any active nodes. The steps are as follows

Double-click Setup.exe set from any SQL server Installation setup media.
The SQL Server Installation Center will be launched by the Installation Wizard.
To add a node to the cluster instance, click Installation in from the pane. Then, opt for Add node option to add a node to SQL Server failover cluster.
The configuration of a new node to the cluster provides option to select language, asks for product key, license terms, server configurations, system configurations, etc.
After installation is Complete the page provides a link to the log file for the installation to the process, click Close.

Sometimes you may be instructed to restart the computer. If prompted then do so. It is important to read all message from Installation Setup.

Remove a Node

In order to remove an existing node from a failover cluster, one must run SQL Server Setup on the node that is to be removed from instance. The process is as follows

Insert the SQL Server installation media.
Double-click setup.exe from the root folder
From The Installation Wizard launched select option to remove a node to an existing failover cluster instance, click Maintenance then select Remove node from a SQL Server cluster.
A discovery operation runs on the computer by the System Configuration Checker Click OK to continue.
System Configuration Checker verifies state of your system and then continues to next.
Click Remove to continue the removal process.
To completion of the removal of a node from SQL Server cluster, click Close.

Rather than the addition and removal of the nodes to a SQL server Failover cluster, upgrading, renaming, client tool installations, etc. can be carried out in the cluster instances.

Bottom Line

SQL Server Failover Clustering is an advanced technique maintained to have high data availability. Even though it costs higher, as far as the value of data is concerned, most of the organizations maintains SQL Server failover clustering. This saves the organization from downtime.