Database Components Cluster Configuration

Plan and prepare the configuration according to the information set out in this appendix. The appendix contains information specific to BMC-supplied PostgreSQL and third-party database servers (for example, Oracle, MSSQL) that are dedicated to a Control-M application.

If you are planning for clusters on a third-party database server that is not dedicated to a Control-M application, contact the site DBA and Control-M administrator regarding specific cluster requirements.

In a PostgreSQL or existing third-party database server installation, the database server is always installed within the same file system, and the Control-M tablespace can reside on a different file system. Both file systems must belong to the same resource group.

You can use the following types of implementations for the Control-M database in a Windows cluster environment:

• clustered configuration: A single database installation on the cluster can be accessed by both nodes. The database files are located on the shared disk and the database server availability is controlled by the cluster.

• remote database configuration: The database is installed on a non-clustered server or on a cluster different from where Control-M is installed. In this case, only database clients are installed on each of the Control-M cluster nodes.

The database server and the Control-M database must be located on the same file system as Control-M.

Microsoft SQL (MSSQL) Server

When MSSQL is installed on a cluster, the software binaries of the product are placed on the local drives of each one of the selected nodes and the data files are placed on the assigned disk resource on the shared drive.

The MSSQL Server services (one service is created on each cluster node) are named by default as the virtual server name. When the MSSQL resource group is moved from one node to another, the MSSQL services are stopped on the original (primary) node and started on the alternate node.

In addition to the MSSQL services resources, installation of MSSQL Server on a Windows cluster automatically creates the network name and IP address resources in the selected resource group. At that point, the MSSQL resource group becomes a virtual server and can be accessed by client applications using the virtual name.

Oracle Real Application Cluster (RAC)

A normal Oracle installation consists of a single Oracle instance that accesses a database on the same computer system. With RAC (formerly known as Oracle Parallel Server), multiple instances on different nodes can access the same database files simultaneously. In case of a node failure, the workload of this node will be handled by the other node of the cluster.

When the system is prepared for installation, the Oracle Universal Installer (OUI) presents the list of all cluster nodes, enabling the selection of a subset as targets. The OUI copies the Oracle software onto the first node, and then propagates the software onto the rest of the chosen nodes of the cluster. Along with all the Oracle software, Oracle Enterprise Manager (the central managing console) is automatically installed and set up. When the installation is finished, the database creation wizard and the network configuration wizard are automatically invoked.

Oracle high availability features are only relevant for Control-M/Enterprise Manager.

Remote Database Server Configuration

When the database server is located on a remote node (not one of the cluster nodes), a database client can be installed on each of the cluster nodes and which can then be connected to the remote database server. In this type of implementation, except for the database client, additional database software does not need to be installed on the cluster.

This configuration is common when database platforms reside in a centralized location. However, the availability of Control‑M in this type of configuration is tied to and dependent upon the remote database server availability.