HP XC System Software Installation Guide Version 4.0

13 Adding Visualization Nodes to an Existing HP XC System
This chapter describes how to install and configure visualization nodes into an existing HP XC
system after the SVA nodes have been fully integrated into the hardware configuration. It
addresses the following topics:
• “Prerequisites” (page 171)
• “Installation Scenarios” (page 171)
13.1 Prerequisites
Before you begin, ensure that the HP XC system meets the following prerequisites:
• The HP XC system is fully installed and configured
• New visualization nodes and support hardware have been added (cabled) to the system
• With the exception of the head node, the power on all nodes is turned off
13.2 Installation Scenarios
Several installation scenarios are possible. Determine the scenario that applies to your specific
hardware configuration and proceed to the appropriate section. The installation procedure you
follow depends on the number of nodes you previously supplied to the cluster_prep command.
• “New Visualization Nodes Exceed the Maximum Number of Nodes Supplied to the
cluster_prep Command” (page 171)
• “New Visualization Nodes Do Not Exceed the Maximum Number of Nodes Supplied to the
cluster_prep Command” (page 172)
• “Graphics Cards Have Been Added to Existing Nodes” (page 172)
13.2.1 New Visualization Nodes Exceed the Maximum Number of Nodes Supplied
to the cluster_prep Command
Follow this procedure if you added new SVA visualization nodes to the hardware configuration,
and by adding the new nodes exeeded the maximum number of nodes set during the initial
cluster_prep process.
Exceeding the previous maximum number of nodes means that you must run the cluster_prep
command again to specify the new maximum number of nodes in the hardware configuration
and to discover the new nodes.
1. Follow the instructions in Section 2.4.2 (page 48) to install the SVA software.
2. Reset the database:
# /opt/hptc/sbin/reset_db
3. Start with Chapter 3 (page 51), and follow the instructions to run the cluster_prep,
discover, and cluster_config utilities.
4. Set all nodes to network boot:
# setnode --resync --all
5. See Section 6.7 (page 117), and follow the instructions to run the startsys utility to reimage
all nodes.
6. Perform the system verification tasks in Chapter 8 (page 129).
13.1 Prerequisites 171