Description of problem: The setting of "path" line on /etc/kdump.conf is not reflected. The dumpfile is always saved to "/var/crash". Please fix to reflect the setting of "path" line by correcting the "/etc/init.d/kdump" script. Version-Release number of selected component (if applicable): kexec-tools: RHEL5.0 - RHEL5.1 (1.101-194.4.el5) How reproducible: kernel crashdump Steps to Reproduce: /etc/kdump.conf: ... path /tmp/ ... Actual results: It created vmcore into /var/crash/127.0.0.1-xxxxxx. Expected results: It created vmcore into /tmp/ (arbitrary directory). Please fix the kdump script.
If "ext3" line is "another partition (ex: /dev/cciss/c0d0p6)" and "path" line is "/crash1", I confirmed that I could get a dumpfile. If "ext3" line is "rootfs partition (ex: /dev/cciss/c0d0p1)" and "path" line is "/crash2", I could get a dumpfile. If "ext3" line is "not set" and "path" line is "/crash3", I could not get a dumpfile. The system causes "kernel panic on 2nd kdump kernel" or "reboot without get a dumpfile on 2nd kdump kernel". However, it is not possible to set it by "system-config-kdump" like it. Therefore, its problem is not a major problem. Please close this issue.
Created attachment 432214 [details] enable path option for kdump Hi, I have also encountered the problem described in the original comment, but it is really easy to fix by inserting a few lines in kdump.init. I have attached my proposed patch. It tests and works on my system. Test case: /etc/kdump.conf: ... path /var/dump ... echo c > /proc/sysrq-trigger Actual results: Full, readable core created in /var/dumps/2010-07-14-18\:14/vmcore
Sarina, your patch looks fine for me, I will take it. Thanks!
Verified with -107.el5: ============================================================ [root@hp-xw4550-01 ~]# rpm -q kexec-tools kexec-tools-1.102pre-107.el5 [root@hp-xw4550-01 ~]# tail /etc/kdump.conf #core_collector makedumpfile -c --message-level 1 #core_collector cp --sparse=always #link_delay 60 #kdump_post /var/crash/scripts/kdump-post.sh #extra_bins /usr/bin/lftp #disk_timeout 30 #extra_modules gfs2 #options modulename options #default shell path /var/kdump [root@hp-xw4550-01 ~]# ls /var/ account cache cvs empty gdm local log nis preserve run tmp beah crash db games lib lock mail opt racoon spool yp [root@hp-xw4550-01 ~]# service kdump restart Stopping kdump:[ OK ] Detected change(s) the following file(s): /etc/kdump.conf Rebuilding /boot/initrd-2.6.18-214.el5kdump.img Starting kdump:[ OK ] [root@hp-xw4550-01 ~]# echo c > /proc/sysrq-trigger ------------------------------------------------------------------------------------------------------------------------ [root@hp-xw4550-01 ~]# ls /var/kdump/2010-09-03-02\:23/vmcore /var/kdump/2010-09-03-02:23/vmcore ============================================================ Change status to VERIFIED.
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2011-0061.html