HP XC System Software Administration Guide Version 3.2

See netdump(8) and netdump-server(8) for more information.
7.10.4 Obtaining the Kernel Dump
When a node running the netdump service experiences a kernel crash, its oops message and
kernel memory are automatically transmitted to the netdump-server node.
This data is stored in a subdirectory (identified by date and time) of the /var/crash directory
on the netdump-server node. After the data is saved, the node that crashed reboots.
7.10.5 Using the Crash Utility to Analyze a Kernel Dump File
You can use the crash utility to examine a kernel dump file that was transmitted to a
netdump-server node. You can analyze the dump file from a user account; you do not need
superuser permission to run the crash utility.
Enter the following command on the netdump-server node to analyze a kernel dump file:
# crash mapfile namelist dumpfile
where:
mapfile
The mapfile corresponding to the kernel the system was originally booted from
(before kernel dump) is System.map.
namelist The uncompressed kernel image compiled with -g option is vmlinux.dbg.
dumpfile The name of uncompressed kernel core dump obtained by the netdump utility is
vmcore.
Usually, the command line is as follows:
# crash System.map vmlinux.dbg vmcore
See crash(8) for more information.
7.10.6 Using the Crash Utility to Analyze a Live System
You must log in as superuser (root) to analyze a system while it is running.
Use the following command to start the crash utility:
# crash vmlinux.dbg
NOTE: In this example, the vmlinux.dbg file is the uncompressed kernel image (that is, a
vmlinux file) that has been compiled with the -g option. The vmlinux.dbg file is the same
version as the live system kernel.
See crash(8) for more information.
104 Monitoring the System