Tuesday, March 27, 2012

How to generate a crash dump on Redhat Linux

This is a guide on how to gather a crash dump on RHEL systems.

First, install the following packages:
# yum install kexec-tools crash kernel-debuginfo kernel-debuginfo-common system-config-kdump 

In graphical mode:
# system-config-kdumpIn the target settings tab:
Select the appropiate Path: and Partition Values.
Click Apply.
Reboot the system.

In console mode:
Edit the /boot/grub/menu.lst file:

Add "crashkernel=128M" to the kernel line:
kernel /boot/vmlinuz-2.6.32-54.el6.i686 root=/dev/sda3 ro crashkernel=128M


Modify the config file: /etc/kdump.conf

Modify these values for where the core file should be stored:

ext3 /dev/sda4
path /var/core

#reboot


Capture the core: (This will panic the system and may take a long time to complete)

# echo c > /proc/sysrq-trigger Locate the core:
# ls /var/crash Analyze the core:

# crash /var/crash/2012-03-29/vmcore /usr/lib/debug/lib/modules/`uname -r`/vmlinux

Available Commands:
crash> sys
crash> bt -a
crash> mod
crash> log
crash> sys