Installing and configuring the management subsystem

Use this information to learn about the requirements for installing and configuring the management subsystem.

The management subsystem installation and configuration encompass major tasks M1 through M4 as shown in Figure 1.

This procedure is not a detailed description of how to install the management subsystem components, because such procedures are described in detail in documentation for the individual devices and applications. This procedure documents the order of installation and key points that you need to consider in installing and configuring the management consoles.

This task is the most complex area of a high-performance computing (HPC) cluster installation. It is affected by, and affects, the other areas (such as server installation and switch installation). Many tasks can be performed simultaneously, while others must be done in a specific order.

You need to set up and configure a Network Installation Management (NIM) server that contains a Shared Product Object Tree (SPOT) to run diagnostics for servers without removable media capabilities (CD and DVD drives). The diagnostics are only available with the AIX® operating system and require an AIX NIM SPOT even if logical partitions are running the Linux® operating system.

If your logical partitions are running the Linux operating system, you also need a distribution server for updating the operating system that is to be used on the logical partitions in servers without removable media capabilities.

While it is typical to use the CSM/MS as the Dynamic Host Configuration Protocol (DHCP) server for the service VLAN, if a separate DHCP server is installed, follow the DHCP installation tasks in the Installing the CSM Management Server topic.

The management consoles that are to be installed are the HMC, the CSM/MS, and the fabric management server. The management consoles are the key to successfully installing and configuring the cluster because they are central to the management subsystem. Before you perform any startup and configuration, these devices must be installed and configured so that they are ready to discover and manage the rest of the devices in the cluster.

During management subsystem installation and configuration, perform the following tasks, which are illustrated in Figure 1. While many of the steps within the procedures can be performed simultaneously, pay attention to where they converge and where one task might be a prerequisite for another task. When there is a prerequisite, a triangle symbol is shown. The following steps are the major tasks for installation:

1. Physically place units on the data center floor.
2. Install and configure service and cluster VLAN devices by using the procedure in Installing and configuring service VLAN devices.
3. Install the HMCs. See Installing the Hardware Management Console.
4. Install the CSM Management Server. See Installing the CSM Management Server.
5. Install the operating system installation servers. See Installing operating system installation servers.
6. Install the fabric management server. See Installing the fabric management server.
7. Perform server installation and configuration with management consoles. See Installing and configuring the cluster server hardware.
8. Configure remote logging from the switches and the fabric management servers to the CSM/MS. See Setting up remote logging.
9. Configure the remote command processing capability from the CSM/MS to the switches and fabric management servers. See Setting up remote command processing.

The tasks have reference labels to use as cross-references between figures and procedures. The first is from Figure 1 and the second is from Figure 1. For example, E1 (M1) indicates task label E1 in the following figure and task label M1 in the Figure 1 topic.

Steps that have a shaded background are steps that are performed under Installing and configuring vendor InfiniBand switches.

Figure 1. Management subsystem installation tasks