HP XC System Software Installation Guide Version 3.1

6 Reinstalling Version 3.1
This appendix describes how to reinstall HP XC System Software Version 3.1 on a system that is already
running Version 3.1.
Reinstalling an HP XC system with the same release may be necessary if you participated as a field test
site of an advance development kit (ADK) or an early release candidate kit (RC).
The procedure is different depending on the hardware model:
• “Reinstall Systems with HP ProLiant Hardware Models” (page 109)
• “Reinstall Systems with HP Integrity Hardware Models” (page 110)
6.1 Reinstall Systems with HP ProLiant Hardware Models
This section describes the following tasks:
• “Reinstall the Entire System” (page 109)
• “Reinstall One or More Nodes” (page 109)
6.1.1 Reinstall the Entire System
Follow this procedure to reinstall HP XC System Software Version 3.1 on systems containing HP ProLiant
hardware with Opteron or Xeon chip architectures:
1. Begin this procedure as the root user on the head node.
2. If SLURM is configured, ensure that all jobs are drained from all nodes, where nodelist represents
the entire set of client nodes (for example, n[1–64]):
# scontrol update NodeName=nodelist State=DRAIN Reason="system shutdown"
3. Shut down all nodes:
# stopsys
4. Insert the HP XC Version 3.1 DVD distribution media in the DVD tray on the head node.
5. Reboot the head node to start the Kickstart installation:
# reboot
6. Follow the instructions in this document, beginning with Chapter 2 (page 33), to complete the system
installation.
6.1.2 Reinstall One or More Nodes
Follow this procedure to reinstall one or more HP ProLiant client nodes from the latest golden image on
the image server. This procedure sets the specified nodes to network boot from the DHCP server and
automatically starts a reinstallation operation on the specified nodes by temporarily removing the
appropriate local boot file in the /tftpboot/pxelinux.cfg directory for each node to be reinstalled.
This example assumes nodes n1 through n5 are being reinstalled.
1. Begin this procedure as the root user on the head node.
2. If SLURM is configured, ensure that all jobs are drained from the affected nodes:
# scontrol update NodeName=n[1-5] State=DRAIN Reason="system shutdown"
3. Prepare the specified nodes to network boot instead of booting from local disk:
# setnode --resync n[1-5]
4. Reboot nodes n1 through n5:
6.1 Reinstall Systems with HP ProLiant Hardware Models 109