Hi,
I am trying to capture a kernel crash dump to help debugging hard freezes. I have the following services enabled and running:
- kexec-load
- kdump
- kdump-early
- kdump-commandline
The journal indicates that the crash kernel is configured
journalctl -b -1 | grep -E '(kdump|crash)'
Jul 29 15:25:57 HOSTNAME kernel: Command line: root=/dev/mapper/system-root splash=silent mitigations=auto quiet security=apparmor crashkernel=72M,low crashkernel=428M,high
Jul 29 15:25:57 HOSTNAME kernel: crashkernel low memory reserved: 0x6c000000 - 0x70800000 (72 MB)
Jul 29 15:25:57 HOSTNAME kernel: crashkernel reserved: 0x000000075a000000 - 0x0000000774c00000 (428 MB)
Jul 29 15:25:57 HOSTNAME kernel: Kernel command line: root=/dev/mapper/system-root splash=silent mitigations=auto quiet security=apparmor crashkernel=72M,low crashkernel=428M,high
Jul 29 15:25:58 HOSTNAME dracut-cmdline[361]: Using kernel command line parameters: rd.driver.pre=btrfs rd.luks.uuid=luks-8695af63-bbc1-443d-9cb1-b37c52222910 rd.lvm.lv=system/root rd.lvm.lv=system/swap resume=UUID=71535983-7598-46f3-8193-5dd7572237e1 root=UUID=2e1135f1-78ed-439a-abd4-93f431981a19 rootfstype=btrfs rootflags=rw,relatime,ssd,space_cache=v2,subvolid=266,subvol=/@/.snapshots/1/snapshot,subvol=@/.snapshots/1/snapshot root=/dev/mapper/system-root splash=silent mitigations=auto quiet security=apparmor crashkernel=72M,low crashkernel=428M,high
Jul 29 15:25:58 HOSTNAME plymouthd[601]: 00:00:01.790 ply-utils.c:959:ply_get_kernel_command_line : Kernel command line is: 'root=/dev/mapper/system-root splash=silent mitigations=auto quiet security=apparmor crashkernel=72M,low crashkernel=428M,high
Jul 29 15:26:11 HOSTNAME plymouthd[1417]: 00:00:14.380 ply-utils.c:959:ply_get_kernel_command_line : Kernel command line is: 'root=/dev/mapper/system-root splash=silent mitigations=auto quiet security=apparmor crashkernel=72M,low crashkernel=428M,high
Jul 29 15:26:12 HOSTNAME systemd[1]: Starting Load kdump kernel early on startup...
Jul 29 15:26:13 HOSTNAME systemd[1]: Finished Load kdump kernel early on startup.
Jul 29 15:26:13 HOSTNAME systemd[1]: Started Check and update kdump options on the kernel command line.
Jul 29 15:26:13 HOSTNAME kdumptool[1901]: Kdump expects these crashkernel= values on the kernel command line:
Jul 29 15:26:13 HOSTNAME kdumptool[1901]: crashkernel=72M,low crashkernel=419M,high
Jul 29 15:26:13 HOSTNAME kdumptool[1901]: (based on the output of 'kdumptool calibrate')
Jul 29 15:26:13 HOSTNAME kdumptool[1901]: /proc/cmdline contains:
Jul 29 15:26:13 HOSTNAME kdumptool[1901]: crashkernel=72M,low crashkernel=428M,high
Jul 29 15:26:13 HOSTNAME kdumptool[1901]: Kdump may not work correctly
Jul 29 15:26:14 HOSTNAME systemd[1]: Started Send e-mail notifications for new crash dumps.
Jul 29 15:26:14 HOSTNAME systemd[1]: Starting Load kdump kernel and initrd...
Jul 29 15:26:14 HOSTNAME kdump-notify[2455]: KDUMP_NOTIFICATION_TO not configured
Jul 29 15:26:14 HOSTNAME systemd[1]: kdump-notify.service: Deactivated successfully.
Jul 29 15:26:14 HOSTNAME systemd[1]: Finished Load kdump kernel and initrd.
However, when I trigger a kernel crash manually there is no crash information captured and no reboot happens. I waited 5 minutes last time.
I set KDUMP_VERBOSE=16 in /etc/sysconfig/kdump but this adds no extra information.
How can I find out why kdump does not save crash information?