In the above illustration, there are two logical data drives in the shared disk array I can select, either F: or G:. I can only select one at this time, and I have selected G:. Later, once SQL Server has been installed, you can also specify that the SQL Server 2000 cluster can use other logical drives on the shared disk array. But at this time, I can only select one logical disk drive.
Essentially, what this step does is to tell the setup wizard where to install the SQL Server 2000 system database files. In addition, it tells the setup wizard that this disk resource will become part of the SQL Server clustering resources. This ensures that if SQL Server should failover, that the disk resource used by SQL Server for the system databases will also fail over with the rest of the SQL Server resources as a single unit. Once you are done, click “Next,” and this screen appears:
8) This screen is used to tell the setup wizard which nodes SQL Server should be installed on. Assuming this is a two-node cluster, the names of both of the two nodes should be in the “Configured Nodes” box. If not, then you can click on the “Add” button to move them from the “Available Nodes” box to the “Configured Nodes” box. When you are ready, click “Next,” and the following screen appears:
9) This screen needs some explanation, as it is not obvious what it is for. When the setup wizard installs SQL Server 2000, it will install it on the local drive of both nodes of the cluster, along with the disk array selected in the previous step. One of these nodes, the primary node, is the server where the wizard is currently running. The other node is the called the secondary node.
To install SQL Server 2000 on the secondary node from the primary node, a remote SQL Server 2000 installation is performed. In order for a remote installation to work, the wizard must login to the secondary node as an administrator.
This is the purpose of this screen, to enter a login name and password of an account that is a local administrator of both nodes in the cluster. In the above illustration, I have used the SQL Server service account domain name. I could have used any account that is a local administrator on both nodes of the cluster, but I chose this one because I know it is already a member of the local administrators group of both nodes. Once you enter the information, click “Next,” and this screen appears:
10) This screen is used to tell the setup wizard the name of the instance of SQL Server 2000 that are are installing. Since this is the first instance of SQL Server on the cluster, this will be a “default” installation. A named instances is only required if you intend to run more than one instance of SQL Server on the cluster, which is not recommended from a performance point of view. Click “Next,” and this screen appears: