Hi folks
This is the third and the last part of our article series about MS SQL Server 2008 R2 clustering and untill now we completed the creation of virtual machines and Windows clustering on them. In this last part, I will install MS SQL Server 2008 R2 on these virtual clustered machines. But first of all, I want to talk a little bit about MS SQL Server clustering options.
With the release of MS SQL Server 2008 product (I refer it as SQL throughout the article), Microsoft introduced two installation options for it on a clustered environment. These are; Integrated installation and Advanced/Enterprise installation. Integrated installation satisfies the most common requirements for SQL Server deployments. It is used for creating a SQL failover instance, adding a node to an existing cluster, and removing a node from an existing cluster regardless of the original installation option. On the other hand, the Advanced/Enterprise installation option enables users to take advantage of a two-phase SQL Server failover cluster installation process. This is beneficial in enterprise deployment solutions that use technologies such as Systems Management Server (SMS) or scripted deployments. Table 1 represents the installation phases of the two methods.
Table 1: Installation phases
| Method | Phases |
| Integrated installation |
|
| Advanced/Enterprise installation |
Note: The Prepare step does not require an underlying Windows Server failover cluster to be established. The Complete step, however, does require that the underlying Windows Server failover cluster exists. If it does not, setup provides an error and exit. |
In my scenario, I will prefer the first method which is Integrated Installation. I will first install SQL on one of my servers as usual which will create an SQL cluster instance and I will add the second server as a member of that instance. So, don’t waste time and start with the installation of first server.
Because I installed Windows Server 2008 R2 as the operating system, I don’t need any hotfix/service pack requirements to install for SQL deployment. The only thing that I will install before SQL installation is MS DTC (Microsoft Distributed Transaction Coordinator). Actually this is not a necessity but for active/active SQL failover clustering, MS DTC is a must. You can find details on http://technet.microsoft.com/en-us/library/cc730992(WS.10).aspx webpage.
Note: If you try to install SQL without MS DTC, you will receive a warning message like Figure 1. But that doesn’t mean you could not install SQL without a clustered MS DTC service.
Figure 1: Warning message about MS DTC
As you remember from the first part of the article, I created 3 shared disks for clustering and one of them is for MS DTC (the other two is for quorum and SQL). So, for installation of MS DTC in a clustered environment I log on one of my servers and open the “Failover Cluster Manager” snap-in. I right click the server name and select “Configure a Service or Application…”. This opens a wizard and first screen is an informative screen as usual. After I press Next button, I select “Distributed Transaction Coordinator (DTC)” from the list (Figure 2) and press Next button. In this screen, I give a name of “mstipclusterdtc” and an IP address of 192.168.2.103/24 for my new clustered service (Figure 3) and press Next. On the coming screen, I select one of my shared storages for MS DTC (Figure 4) and press Next. After the Confirmation and Summary screens, the wizard is finished and I have MS DTC service in a clustered environment.
Figure 2: “Distributed Transaction Coordinator” selection
Figure 3: Name and IP for MS DTC service
Figure 4: Storage for MS DTC Service
So it is SQL Server installation time :). For this purpose, I mount the ISO image of SQL installation DVD to CD-Drive of my first virtual server and run the setup.exe file. It opens a screen for me and I select Installation on the left side. Installation selection offers me five different links and one of them is “New SQL Server failover cluster installation” (Figure 5). I select it and an installation wizard appears.
Figure 5: Installation type selection
The first three-four screens are listing the problems/requirements found in the environment that SQL installation will be performed. In my case, there is no problem and all the items are signed as green (Figure 6).
Figure 6: No problem at all :)
The next two screens are Product Key and License Term screens. After these, I fill all the checkboxes in Feature Selection screen as it is seen in Figure 7 (of course you have to choose whatever you need). In Instance Configuration screen, I give a name to my SQL cluster (Figure 8) and press Next button. After checking Disk Space Requirements screen, Cluster Resource Group screen appears. Here in this screen, some of the previously created resource groups are listed but they are all used by other clustered services. Therefore, to create a new resource group I directly press the Next button here (Figure 9).
Figure 7: “Feature selection” screen
Figure 8: “Instance Configuration” screen
Figure 9: New Cluster Resource Group creation for SQL Server
“Cluster Disk Selection” is the screen that we assign a shared disk to our SQL cluster. After that “Cluster Network Configuration” screen appears. I give an IP address of 192.168.2.104 for my SQL cluster here (Figure 10). I select the default and recommended selection (use service SIDs) in “Cluster Server Policy” screen. In “Server Configuration” screen, service accounts and collation types are selected/entered. After that “Database Engine Configuration” and “Analysis Services Configuration” screens appear. In these parts of the setup, we define the administrative user accounts for database engine and analysis services. Also database locations on shared disk are defined here (Figure 11). Let me remind you that, these screen appearances can be different in your environment if you select less feature than me ( I selected all the features at the beginning of the setup wizard, if you remember). The rest of the screens are informative or checking screens and by clicking Next button on each of them I finished the installation of SQL on my first machine successfully (Figure 12).
Figure 10: “Cluster Network Configuration” screen
Figure 11: Database location of SQL server
Figure 12: Wizard completion screen
Now it is time to add the second node to my SQL cluster. To do this, I logon to the second virtual machine and run SQL setup there. I select “Add node to a SQL Server failover cluster” on the opening screen (Figure 13). The wizard is almost the same but a short version. I mean that the first 4-5 screens are the same but after “Setup Support Rules” screen, “Cluster Node Configuration” screen appears (Figure 14) and I select my one and only SQL Cluster instance in that screen and finish the wizard by pressing Next buttons on the following screens.
Figure 13: “Add node to a SQL Server failover cluster” selection
Figure 14: “Cluster Node Configuration” screen
As I finished the SQL installation in a clustered environment and add the second node to my cluster instance, I have to check whether it works or not. I open the “Failover Cluster Manager” snap-in and under “Services and Applications” I click SQL Server. The screen summarizes the condition of SQL service (Figure 15). To check whether the second cluster node handle the job when the first node fails, you can restart the first machine and control the“Failover Cluster Manager” snap-in on the second virtual machine. You have to see that owner of SQL services changes to the second node.
Figure 15: SQL Server summary screen
This is the end of my three part article series. I hope it would be helpful to you. See you next time.
RSS feed for comments on this post.
