HP XC System Software Installation Guide Version 4.0

• The Portland Group, supplier of the C/C++ and Fortran compilers
http://www.pgroup.com/
5.9 Creating Space for Kernel Crash Dumps (Optional)
This task is optional. If you want to configure the kdump facility to save and store kernel crash
dumps, you must create the space to store them. The cluster_config utility will prompt you
to enter the mount point for storing the crash dumps.
IMPORTANT: Be aware that if you do not configure the kdump facility, no other mechanism is
available to save and store kernel crash dumps.
The netdump utility, which was a supported crash dump tool, is no longer supported for use
by the underlying Linux base OS.
If you elect not to make the kdump facility highly available, the crash dump area can be a file
system or a directory in a file system. HP recommends creating a dedicated file system for this
purpose. You must mount this file system on the head node before you run the cluster_config
utility.
IMPORTANT: Do not use an area on the root or /hptc_cluster file systems for the crash
dump area because if the crash dump area becomes full, cluster performance or node activities
might be impacted.
The crash dump area must be large enough to store the number of crash dumps you want to
keep.
If you are configuring the kdump facility to be highly available, the crash dump area must be a
separate Logical Volume on shared fiber channel storage. In an improved availability
configuration, the kdump server is relocated to the backup server when the head node is
unavailable.
Use the instructions in Section 3.6.3 (page 61)to create and mount an LVM volume for the crash
dump area (the steps are the same as creating and mounting an LVM volume for the
/hptc_cluster file system).
5.10 Modifying Workstation Model Names in the Database (Optional)
This task is optional. Perform this task if the hardware configuration contains HP xw8400 or
xw9400 workstations. Omit this task if the hardware configuration does not contain these models
of HP workstations.
Follow this procedure to correctly identify HP xw8400 or xw9400 workstations in the configuration
and management database:
1. Begin this procedure as the root user on the head node.
2. Change to the following directory:
# cd /opt/hptc/config/sbin
3. Modify the workstation model name in the database. The node_list in the command
identifies the nodes that are the workstations. You can determine the workstation node
names based on their cable position on the Administration switch.
Specify the node_list in the format, node_prefix[n-n], where n is the node number.
For example, prefix[1-3,7,8].
The command depends on the workstation model:
• Enter the following command to change the name of xw8400 workstations:
# ./modify_node_type "hp workstation xw8200" \
"HP xw8400 Workstation" node_list
88 Setting Up the Environment