If the system does not restart after the memory dump is
100% completed, restart the system manually.
After collecting memory dump
Check that the memory dump is collected after the system re-starts. The default
registry for memory dump is "/var/crash/<date-time>/". <date-time> is when the
memory dump is collected.
If the memory dump file does not exist in the default location
after memory dump is collected, the location may be
changed from the default. Check the location written on the
setting file.
Setting file name: /etc/kdump.conf
Memory dump for VMware
This subsection describes how to collect memory dump for Linux.
See "Using Parameters to Stop the System at Hardware
Failure" and "Useful Settings for VMware" in Chapter 13 for
memory dump settings.
Triggering memory dump
When the local console displays a purple screen, memory dump is automatically
collected. When the system hangs up, issuing NMI allows the memory dump to be
collected. Follow the steps below to issue NMI.
1
Register the console screen.
2
Check if the virtual machine is running via the network from another machine. If the
virtual machine is running, shut it down.
3
Issue NMI.
To issue NMI, select NMI from the Power and Reset menu of Remote console
application, or click Server Operation tab > Power and LEDs, and click NMI. See
"Connect the Remote Console" in Chapter 3 or Chapter 5, Server Blade Setup for
details.
1822