I had the need to test a solution against a SQL Server 2008 R2 Cluster with the limitation of only have virtual machines to work with. So I started off doing research to find the information on what it would take to build a SQL Server 2008 R2 Failover Cluster. What I did notice is I did not find a blog, doc or Article with all the information I needed or required to get this design developed. The documents I read assumed allot of settings are in place. So as I when thru KB’s and Articles solving one issue, I hit another issue and had to go find a resolution for the next problem.
So this is my attempt to bring all required information together to build this Step By Step Guide on building a 100% Virtualized SQL Server 2008 R2 Cluster for SCCM 2007 R2 to enable High Availability for Self Service or any service application you have a need to provide high availability for.
After building 3 Windows Server 2008 R2 virtual machines with the intent on using 2 as the nodes to the SQL Clustering Service and 1 for the virtual machine acting as the back end storage. I came to realize that you cannot use a virtual Windows Server 2008 R2 machine to host the SCSI drives, this server must be physical.
Crap! Now what I’m going to do. I only have access to a virtual Hyper-V environment…
Back to doing some more research: As it turns out you can use a virtual machine as the back end storage server, but it just cannot be done with the Windows Server 2008 R2 version. It can only be done with the Windows Storage Server 2008 R2 SKU.
Nice! Didn’t know they had this SKU…
So I’m off to search for Windows Storage Server 2008 R2 download in my MSDN subscription. What I found on MSDN was 2 versions. Windows Storage Server 2008 R2 Essentials version and a Embedded version for Standard, Enterprise and Workgroup with iSCSI Software Target 3.3 (x64).
I didn’t read the difference of the two server versions, so I didn’t know how to get storage server installed. Truthfully this is my first time using an embedded application. It took me the whole day of research to determine on how to install Windows Storage Server 2008 R2 before finally realizing that it’s an Embedded version. And there is just a bunch of updates that needs to be installed 1 by 1 that converts the Windows Server 2008 R2 version into Windows Storage Server 2008 R2. After installing the updates and rebooting the server, the server is now Windows Storage Server 2008 R2 SKU. See the difference below.
Before  | After  |
That was a pain! I would script this part if I had many deployments like this to do.
So Let’s get off into the details while I tell this story and provide the Steps to building this solution.
Note: These steps uses Hyper-V for the virtualized environment.
Requirements.
- 3.5.1 .NET Framework Feature
Step 1. Build 3 Virtual Machines.
I’m using Hyper-V in Windows Server 2008 R2 for my virtualized environment. Build 3 virtual machines using Hyper-V in Windows Server 2008 R2. The following list is how I will be using each virtual machines.
- VM1, host installation of SQL Server 2008 R2 for node 1 and Clustering Service.
- VM2, host installation of SQL Server 2008 R2 for node 2 and Clustering Service.
- VM3, host installation of Storage Server 2008 R2 Embedded, SCCM 2007 R3 and SCCM Self Service.
The Virtual Machine configuration settings to use for VM1 and VM2 is outlined below.
- Memory:
- 2048 MB
- CPU:
- 1 Virtual Processor
- Drives:
- 1 IDE Controllers
- 50 GB VHD file for C:\ OS installation
- Virtual DVD Drive
- 1 SCSI Controller
- 50 GB VHD file for I:\ SQL Instance installation
- 1 IDE Controllers
- Network:
- 1 Virtual Network Adapter
The Virtual Machine configuration settings to use for VM3 is outlined below.
- Memory:
- 2048 MB
- CPU:
- 1 Virtual Processor
- Drives:
- 1 IDE Controllers
- 50 GB VHD file for C:\ OS installation
- Virtual DVD Drive
- 1 SCSI Controller
- 100 GB VHD file for E:\ To host the Cluster iSCSI Target shared drives VHD files
- 1 IDE Controllers
- Network:
- 1 Virtual Network Adapter
Step 2. Install & Prepare Windows Server Operating Systems
After you build the 3 virtual machines, Install Windows Server 2008 R2 on all three virtual machines. You will run thru additional steps to convert VM3 from Windows Server 2008 R2 into Windows Storage Server 2008 R2.
Next add each VM to a domain and give each virtual machines virtual adapter static IP address.
Important: Ensure to APPLY ALL REQUIRED UPDATES on all virtual machines. Remember after installing detected updates, run a Windows Update scan after each installation of updates to ensure you have installed all required updates. Note: Some Software Updates will not show as required until other required updates are installed.
Step 3. Setup Windows Storage Server 2008 R2
After you have all 3 virtual machines installed with Windows Server 2008 R2, each assigned with static IP address, you will now configure one of the VM’s (VM3) to become Windows Storage Server 2008 R2.
Download from MSDN the following Windows Storage Server 2008 R2 SKU.
- Windows Storage Server 2008 R2 Embedded (Standard, Enterprise, Workgroup) with iSCSI Software Target 3.3 (x64) - DVD (Multiple Languages)
Note: There is no Product Key required for this installation of Windows Storage Server 2008 R2. This is because it will turn your existing installation of Windows Server 2008 R2 into Storage Server 2008 R2 and uses the existing license.
After you download the ISO file and mount it to your computer or burn it to a DVD, you should see the following file.
- WSS2008R2+iSCSITarget33.exe
Run this EXE to extract the Windows Storage Server 2008 R2 components with elevated rights.
Click Accept on the following screen.
Enter a destination where you want the files to be extracted. Then click Install.
monitor the installation
After extracting the components, you should see the following 2 ISO files.
- Windows_Storage_Server_2008_R2.iso
- iSCSI_Software_Target_33.iso
Mount the (Windows_Storage_Server_2008_R2.iso) file to your computer or burn it to a DVD. You should see the following folder structure.
Open the Windows Storage Server 2008 R2 folder, you should see the following patches.
Install Windows Server 2008 R2 updates
Included with the Windows Storage Server 2008 R2 Component Pack is a collection of Windows Server 2008 R2 updates that must be installed as part of your deployment process. These updates are designed to increase the reliability or security of your storage appliance.
You can deploy these updates in one of three ways:
- By adding updates to the Install.wim (which is included on the product DVD or the ISO file). For more information on how to do this, see the topic “Scenario: Service an Offline Image” in the Windows OPK User's Guide (Opk.chm).
- Use the expand.exe tool to extract all of the .cab files manually from the .msu files (for example, expand *.msu c:\temp), and then use the Deployment Image Servicing and Management (DISM) tool to install all of the .cab files without necessitating system reboots. For more information on the Expand command, see Expand Command-Line Options (http://go.microsoft.com/fwlink/?LinkId=198216). For more information about the DISM tool, see the “Deployment Image Servicing and Management Technical Reference” topic in the Windows OPK User's Guide (Opk.chm).
- Use Windows Update Standalone (WUSA) installer to install updates manually in audit mode. For more information on how to do this, see the “Service an Online Image” topic in the Windows OPK User's Guide (Opk.chm).
- Add references to the updates to an answer file that supplements installation of the Windows Server features when Windows Setup runs.
Apply the following high lighted patches on VM3 to start preparing it to become a Storage Server. Install the patches in the order listed below.
In my environment I’m running Enterprise Server 2008, so in my case I would need to apply the Windows6.1-KB982050-x64-EnterpriseBranding.MSU patch.
- Windows6.1-KB982050-x64-EnterpriseBranding.MSU
- Windows6.1-KB976835-x64-WebRDP.msu
- Windows6.1-KB976833-x64-SIS.msu
You might be asked for a reboot after the last patch is installed. Reboot before continuing anyway.
After applying the patch updates, ensure to reboot VM3. Rebooting the server will complete the installation and provisioning of the Storage Server embedded installation. When the server returns from the reboot, the server will be a Windows Storage Server 2008 R2. You can check this by Right clicking on the Computer and clicking Properties. Shown below.
Now that you have Windows Storage Server running, you now need to install the Microsoft iSCSI Target software 3.3.
Steps 3. Install iSCSC Target Software 3.3
Now mount the (iSCSI_Software_Target_33.iso) ISO file. You should see the following files and folder structure.
Navigate to the folder matching the architecture you are using. In my case I would use the x64 folder. You should see the following files in the folder. Start the installation by running the iscsitarget.msi installer file.
The installation is pretty simplistic, so just run thru the installation wizard and complete the setup.
Now that you have the Microsoft iSCSI Software Target 3.3 software installed, you now need to create simulated virtual shared disks that the SQL Server cluster will use for its shared storage. Virtual disk are created on NTFS volumes as a file. As with any other file, the file system determines where to place the file on the physical disk. This offers flexible provisioning, snapshots that persist after restarting the iSCSI target server, and compatibility with replication software.
.gif)
Note: Don’t worry too much about the size of the virtual disk, they can be extended later. If you have the need to extend you virtual disk, you can find info to accomplish that here: http://technet.microsoft.com/en-us/library/gg232619(WS.10).aspx.
Note: If you set up failover clustering for iSCSI storage, the virtual disk can reside only on a cluster disk.
Configuring a virtual disk for access by more than one iSCSI target can cause data corruption unless the virtual disk is part of a cluster environment or the iSCSI initiators use a storage area network (SAN) file system to coordinate access to the storage.
Install the Failover Clustering Feature.
Now we need to install the Failover Cluster Service on each virtual machine.
Go the the Server Manager and Navigate to the Features Node and Click Add Features.
Click Failover Clustering and then Click Next.
Click Install to start the installation of the Failover Clustering service.
Click Close when the installation of the Failover Clustering Service completes.
Configure the iSCSI Initiators on Node servers
Now you need to configure the initiators on each of the SQL node servers.
Login to one of you servers that will participate in the cluster and go to the Start Menu, Administrative Tools, and Click on Failover Cluster Manager.
REQ: First, Enable Port 3260 in the firewall on all three virtual machines.
The first time you load the iSCSI initiator, it will ask you two questions.
The first question is about loading the Microsoft iSCSI Initiator service every time.
When the iSCSI Initiator windows opens, Click on the Discovery Tab and Click Discover Portal.
Enter the IP address of the Storage Server acting as the target. Click Ok to add the portal.
You should now see something similar to the following.
Configure Failover Clustering Service
Go to the Failover Clustering Console and navigate to the Failover Cluster Manager node and Right Click and Select Create a Cluster.
Click Next on the Before You Begin screen.
On the Select Servers screen, Click the Browse button and enter the name of the servers that will participates as nodes in the cluster.
On the Select Computers dialog and Click Ok to accept the names.
Click Next to continue.
If you see the following screen, you should run the Validation Wizard by leaving the Yes, When I click Next, run configuration validation tests, and then return to the process of creating the cluster. option selected.
Click Next to continue.
Client Next on the Before You Begin screen.
Leave the option Run all tests (recommended) selected and click Next.
Click Next on the Confirmation screen.
Tests will run.
Click Finish when the tests completes to close the Validation Wizard.
Provide a name for the Cluster in the Cluster Name field. This will be the Cluster name you connect to when configuring and managing the SQL cluster.
Provide an IP address that will be assigned to the cluster name. Click Next to continue.
Click Next on the Confirmation screen to start running the Create Cluster process.
The Failover Clustering service will now be created and assign resources.
Click Finish on the Summary screen to complete the process of creating the cluster.
You should now have a Failover Clustering Service running.
Select the Services and applications node, Right Click it and select Configure a Service or Application.
Select the Distributed Transition Coordinator (DTC) service and Click Next to continue.
On the Client Access Point screen, leave the default name given for the name clients will connect to when accessing this service.
Provide an IP address for the network this cluster will be used accessed by.
Click Next to continue.
Select the Cluster Disk to assign to the Distributed Transaction Coordinator (DTC) service and click next to continue.
Click next to create the Distributed Transaction Coordinator (DTC).
Setup and configure the iSCSI Software Target
Ok, now lets get on with configuring the iSCSCI software target.
After the in the installation of the Microsoft iSCSI Software Target, navigate to the Administrative Tools and Click the Microsoft iSCSI Software Target to start the application.
Select the iSCSC Targets node.
Right click the iSCSI Targets node and select Create iSCSC Target.
Click Next on the Welcome to the Create iSCSI Target Wizard screen. Enter a iSCSI target name and Description. Click Next to continue.
On the iSCSI Initiators Identifiers screen, enter the IP address of VM3 and click next.
In the Add iSCSI Inititator, select the iqn you created in
Select the Devices node.
Right click the Devices node and select Create Virtual Disk.
The Create Virtual Disk Wizard will start, Click Next to continue.
You will see the following screen allowing you to the VHD file name and location where the virtual disk will be created. Here you want to select the 100GB VHD drive I created for VM3 the storage server.
Select the Browse button above, then select the drive that is the 100GB drive. In my case the E:\ drive is the 100GB.
After selecting the drive and clicking OK, enter a file name for the VHD virtual disk file. Example shown below. Click Next to continue.
On the Size screen enter the size you want the virtual disk to be created with. Click Next to continue.
On the Description screen enter a description maybe. Click Next to continue.
On the Access screen, click the Add button to Enter a
Select the iSCSI target from the list and click OK.
Click Next to continue.
Then click Finish to complete creating a virtual disk.
Do this three times until you at least have 3 disk with the following sizes for this setup.
- 10 GB
- 30 GB
- 50 GB
Now that the target is created, you can add virtual disks to it. These will be the logical units that will be presented to the initiator. You will do this by right-clicking the target T1 and selecting the option to “Add Existing Virtual Disks to iSCSI Target”.
Click next on the Intro page.
Select the Virtual Disk to add the the cluster Target and click OK.
You should now see all virtual disk you have added in the left pane of the Microsoft iSCSI Software Target console.
Bring the Volumes online
You should now see offline LUNs in the Server Manager under Storage > Disk Management.
Right Click each unallocated space and bring the drive online.
Then Right Click each drive to initialize the dive.
Select the Disk to initialize and click OK to continue.
Now Right Click the drive and Click New Simple Volume.
Ensure you have formatted and assigned a drive letter to each drive.
Setup SQL Cluster
At this point you can follow the installation required for installing a SQL Server instance in a Failover Cluster. You have two option to do this.
- Integration Installation with Add Node
- Advanced / Enterprise Installation
For the instruction this article, I used the Integration Installation with Add Node option. You can find those instruction here: (http://msdn.microsoft.com/en-us/library/ms179530.aspx) under
I pasted those instructions below from the link above. After installing SQL in Cluster mode, you now have a completely virtualized SQL Cluster.
Steps to Install the SQL Cluster. You can access and download the document I used to
Insert the SQL Server installation media, and from the root folder, double-click Setup.exe. To install from a network share, browse to the root folder on the share, and then double-click Setup.exe. For more information about how to install prerequisites, see Before Installing Failover Clustering. You may be asked to install the prerequisites, if they are not previously installed.
Windows Installer 4.5 is required, and may be installed by the Installation Wizard. If you are prompted to restart your computer, restart and then start SQL Server 2008 R2 Setup again.
After prerequisites are installed, the Installation Wizard starts the SQL Server Installation Center. To create a new cluster installation of SQL Server 2008 R2, click New SQL Server failover cluster installation on the installation page
The System Configuration Checker runs a discovery operation on your computer. To continue, click OK. You can view the details on the screen by clicking Show Details, or as an HTML report by clicking View detailed report.
On the Language Selection page, you can specify the language for your instance of SQL Server if you are installing on a localized operating system and the installation media includes language packs for both English and the language corresponding to the operating system. For more information about cross-language support and installation considerations, see Local Language Versions in SQL Server.
To continue, click Next.
On the Setup Support Files page, click Install to install the Setup support files.
The System Configuration Checker verifies the system state of your computer before Setup continues. After the check is complete, click Next to continue. You can view the details on the screen by clicking Show Details, or as an HTML report by clicking View detailed report.
On the Product key page, indicate whether you are installing a free edition of SQL Server, or whether you have a PID key for a production version of the product. For more information, see Editions and Components of SQL Server 2008 R2.
On the License Terms page, read the license agreement, and then select the check box to accept the license terms and conditions. To help improve SQL Server, you can also enable the feature usage option and send reports to Microsoft. Click Next to continue. To end Setup, click Cancel.
On the Feature Selection page, select the components for your installation. A description for each component group appears in the right pane after you select the feature name. You can select any combination of check boxes, but only the Database Engine and Analysis Services support failover clustering. Other selected components will run as a stand-alone feature without failover capability on the current node that you are running Setup on.
You can specify a custom directory for shared components by using the field at the bottom of this page. To change the installation path for shared components, either update the path in the field provided at the bottom of the dialog box, or click the ellipsis button to browse to an installation directory. The default installation path is C:\Program Files\Microsoft SQL Server\.
Note
When you select the Database Engine Services feature, both replication and full-text search are selected automatically. Unselecting any of these sub features also unselects the Database Engine Services feature.
On the Instance Configuration page, specify whether to install a default or a named instance. For more information, see Instance Configuration.
SQL Server Network Name — Specify a network name for the new SQL Server failover cluster. This is the name that is used to identify your failover cluster on the network.
Note
This is known as the virtual SQL Server name in earlier versions of SQL Server failover clusters.
Instance ID — By default, the instance name is used as the Instance ID. This is used to identify installation directories and registry keys for your instance of SQL Server. This is the case for default instances and named instances. For a default instance, the instance name and instance ID would be MSSQLSERVER. To use a nondefault instance ID, select the Instance ID box and provide a value.
Note
Typical stand-alone instances of SQL Server 2008 R2, whether default or named instances, do not use a nondefault value for the Instance ID box.
Instance root directory — By default, the instance root directory is C:\Program Files\Microsoft SQL Server\. To specify a nondefault root directory, use the field provided, or click the ellipsis button to locate an installation folder.
Detected SQL Server instances and features on this computer - The grid shows instances of SQL Server that are on the computer where Setup is running. If a default instance is already installed on the computer, you must install a named instance of SQL Server 2008 R2. Click Next to continue.
The Disk Space Requirements page calculates the required disk space for the features that you specify, and compares requirements to the available disk space on the computer where Setup is running. For more information, see Disk Space Requirements.
Use the Cluster Resource Group page to specify the cluster resource group name where SQL Server virtual server resources will be located. To specify the SQL Server cluster resource group name, you have two options:
Use the drop-down box to specify an existing group to use.
Type the name of a new group to create. Be aware that the name "Available storage" is not a valid group name.
On the Cluster Disk Selection page, select the shared cluster disk resource for your SQL Server failover cluster. The cluster disk is where the SQL Server data will be put. More than one disk can be specified. The Available shared disks grid displays a list of available disks, whether each is qualified as a shared disk, and a description of each disk resource. Click Next to continue.
Note
The first drive is used as the default drive for all databases, but can be changed on the Database Engine or Analysis Services configuration pages.
On the Cluster Network Configuration page, specify the network resources for your failover cluster instance:
Network Settings — Specify the IP type and IP address for your failover cluster instance.
Click Next to continue.
Use this page to specify Cluster Security Policy.
Windows Server 2008 and later versions - Service SIDs (server security IDs) are the recommended and default setting. The option to specify domain groups is available but not recommended. For information about service SIDs functionality on Windows Server 2008, see Setting Up Windows Service Accounts.
On Windows Server 2003, specify domain groups for SQL Server services. All resource permissions are controlled by domain-level groups that include SQL Server service accounts as group members.
Click Next to continue.
Note
If you are installing a SQL Server 2008 R2 Failover Cluster Instance in a Windows 2000 mixed mode domain, you must use domain global groups for SQL Server Clustered Services.
Note
Windows 2000 domain controllers can operate in mixed mode and native mode. Mixed mode enables down-level domain controllers in the same domain.
Work flow for the rest of this topic depends on the features that you have specified for your installation. You might not see all the pages, depending on your selections (Database Engine, Analysis Services, Reporting Services).
On the Server Configuration — Service Accounts page, specify login accounts for SQL Server services. The actual services that are configured on this page depend on the features that you selected to install.
You can assign the same login account to all SQL Server services, or you can configure each service account individually. The startup type is set to manual for all cluster-aware services, including full-text search and SQL Server Agent, and cannot be changed during installation. Microsoft recommends that you configure service accounts individually to provide least privileges for each service, where SQL Server services are granted the minimum permissions they have to have complete their tasks. For more information, see Server Configuration - Service Accounts and Setting Up Windows Service Accounts.
To specify the same logon account for all service accounts in this instance of SQL Server, provide credentials in the fields at the bottom of the page.
Security Note Do not use a blank password. Use a strong password.
When you are finished specifying login information for SQL Server services, click Next.
Use the Server Configuration - Collation tab to specify nondefault collations for the Database Engine and Analysis Services. For more information, see Server Configuration - Collation.
Use the Database Engine Configuration — Account Provisioning page to specify the following:
Security Mode - select Windows Authentication or Mixed Mode Authentication for your instance of SQL Server. If you select Mixed Mode Authentication, you must provide a strong password for the built-in SQL Server system administrator account.
After a device establishes a successful connection to SQL Server, the security mechanism is the same for both Windows Authentication and Mixed Mode. For more information, see Database Engine Configuration - Account Provisioning.
SQL Server Administrators - You must specify at least one system administrator for the instance of SQL Server. To add the account under which SQL Server Setup is running, click Add Current User. To add or remove accounts from the list of system administrators, click Add or Remove, and then edit the list of users, groups, or computers that will have administrator privileges for the instance of SQL Server. For more information, see Database Engine Configuration - Account Provisioning.
When you are finished editing the list, click OK. Verify the list of administrators in the configuration dialog box. When the list is complete, click Next.
Use the Database Engine Configuration - Data Directories page to specify nondefault installation directories. To install to default directories, click Next.
Important
If you specify nondefault installation directories, make sure that the installation folders are unique to this instance of SQL Server. None of the directories in this dialog box should be shared with directories from other instances of SQL Server. The data directories should be located on the shared cluster disk for the failover cluster.
For more information, see Database Engine Configuration - Data Directories.
Use the Database Engine Configuration - FILESTREAM page to enable FILESTREAM for your instance of SQL Server. For more information about FILESTREAM, see Database Engine Configuration - Filestream. Click Next to continue.
Use the Analysis Services Configuration — Account Provisioning page to specify users or accounts that will have administrator permissions for Analysis Services. You must specify at least one system administrator for Analysis Services. To add the account under which SQL Server Setup is running, click Add Current User. To add or remove accounts from the list of system administrators, click Add or Remove, and then edit the list of users, groups, or computers that will have administrator privileges for Analysis Services. For more information, see Analysis Services Configuration - Account Provisioning.
When you are finished editing the list, click OK. Verify the list of administrators in the configuration dialog box. When the list is complete, click Next.
Use the Analysis Services Configuration — Data Directories page to specify nondefault installation directories. To install to default directories, click Next.
Important
If you specify nondefault installation directories, make sure that the installation folders are unique to this instance of SQL Server. None of the directories in this dialog box should be shared with directories from other instances of SQL Server. The data directories should be located on the shared cluster disk for the failover cluster.
For more information, see Analysis Services Configuration - Data Directories.
Use the Reporting Services Configuration page to specify the kind of Reporting Services installation to create. For failover cluster installation, the option is set to Unconfigured Reporting Services installation. You must configure Reporting Services services after you complete the installation.
For more information about Reporting Services configuration modes, see Report Server Installation Options.
On the Error Reporting page, specify the information that you want to send to Microsoft to help improve SQL Server. By default, option for error reporting is disabled. For more information, see Error Reporting.
The System Configuration Checker runs one more set of rules to validate your configuration with the SQL Server features that you have specified.
The Ready to Install page displays a tree view of installation options that were specified during Setup. To continue, click Install.
During installation, the Installation Progress page provides status so that you can monitor installation progress as Setup continues.
After installation, the Complete page provides a link to the summary log file for the installation and other important notes. To complete the SQL Server installation process, click Close.
If you are instructed to restart the computer, do so now. It is important to read the message from the Installation Wizard when you have finished with Setup. For information about Setup log files, see How to: View and Read SQL Server Setup Log Files.
To add nodes to the single-node failover you just created, run Setup on each additional node and follow the steps for AddNode operation. For more information, see How to: Add or Remove Nodes in a SQL Server Failover Cluster (Setup).
Note
If you are adding more than one node, you can use the configuration file to deploy the installations. For more information, see How to: Install SQL Server 2008 R2 Using a Configuration File.
The SQL Server edition you are installing must match across all the nodes in a SQL Server failover cluster. When you add a new node to an existing SQL Server failover cluster, make sure that you specify that the edition matches the edition of the existing failover cluster.
