Managing a Windows and SQL Server Cluster using the Failover Cluster Manager tool

Problem

From a DBA standpoint, I’ve installed my first Windows and SQL Server cluster on the 2008 platform. Now what?

Solution

Installing a cluster is just the beginning of a DBA’s administrative duties. In this tip, I’ll show you a few management tasks that may help the novice. Most tasks will be carried out using Failover Cluster Management which is the management console built into Windows Server 2008. You can access the Failover Cluster Management in a couple of different ways:

  • Go to Server Manager, Features, Failover Cluster Manager
ServerManager
  • Go to Control Panel, Administrative Tools, Failover Cluster Manager
StartBar

Once you open Failover Cluster Manager let’s see what can be done inside of this console.

View Status of Services and Applications

Inside of Failover Cluster Manager, if you click on Services and Applications you are presented with a wealth of information.

In this example, I have a multiple instance cluster that consists of four instances on a two node active/passive cluster. After clicking Services and applications, you can see in the right pane all the instances, status, type, current owner (or current node that they exist on) and if they are set to auto start. In the right pane, if you click on a specific instance you can see more information regarding this instance at the bottom.

SA1

Likewise, if you drilldown from Services and applications and click on an instance name you will see even more information regarding this instance that includes server name, IP address, disk information, and the services that are clustered.

SA2

View Status of Nodes

Click on Nodes to view the status of the nodes in the cluster. Here you will be able to see the nodes that have been installed and the status of each node.

Nodes

View Status of Storage

Click on Storage to view the summary of each storage device. Here you will able to see a total storage summary that includes total capacity, available capacity, and each cluster disk and whether it’s online or offline and the current owner.

Storage

View Status of Networks

Click on Networks to view the status of your cluster networks.

Network1

To view each individual network click on the network desired. Here you will be able to see subnet information, each NIC and the IP addresses associated with them, each NIC status and current owner.

Network2

Perform a Services or Applications Failover

To perform a manual failover, drilldown into services and applications and select an instance name. Right click the instance name, highlight “Move this service or application to another node, and finally click Move to node….

Failover1

In the Please confirm action box, read the message and select the option to Move SQL Server

Failover2

It will show Offline for a few seconds while it is performing the failover, but should soon say Online again and the owner should change.

Failover3

Setting Preferred Owners for Services and Applications

It is possible for you to set each service and application to have a preferred owner at startup meaning when the servers are brought online the service and application will start up under the correct owner.

Below are the steps to manage this option:

Drilldown into Services and Applications, right click an instance name and select Properties.

SAProp1

In the Service and Applications properties box, select the preferred owner you would like the instance to start up in. You can also select all of the owners and move them in order using the Up and Down buttons. (The owner at the top will have top priority)

In this example, I have chosen SQLCL01 and SQLCL02 as my preferred owners, but since SQLCL01 is at the top it will be the owner my instance connects to. You can also view what node this instance is connected to currently at the bottom of this dialog box.

SAProp2

To ensure the service or application automatically moves to the correct node, select Allow Failback under the Failover tab.

SAProp3

Next Steps

  • Another administrative task dealing with clusters is removing nodes which can be viewed in this tip
  • View all of the MSSQLTips Clustering tips here

2 Comments

Leave a Reply

Your email address will not be published. Required fields are marked *