kdump-config(8)

✨ kdump-config(8) ✨

Linux System Administrator's Manual 💅

🎀 NAME

kdump-config — configure kdump, check status, or save a vmcore file 🦋

🔮 SYNOPSIS

kdump-config { help | test | show | status | load | unload | savecore |
      propagate | symlinks {kernel-version} }

💖 DESCRIPTION

okay so kdump-config is the manager for the kdump feature in the Linux kernel 🔥 When a kdump enabled kernel panics, it instantly boots into a clean kernel called the kdump kernel. The memory image of the panicked kernel shows up in /proc/vmcore while the kdump kernel (or "capture kernel") is running. iconic, right? 😭

kdump-config loads the kdump kernel into the current kernel, and after a kdump, it captures the vmcore image through the kdump kernel. It can take advantage of the linux-image-$(uname -r)-dbg package and will try to use makedumpfile to capture the vmcore. Then you can use the crash tool to analyze the vmcore later, no cap 🧚

kdump-config exits with 0 on success and 1 on failure. Simple, clean, periodt ✨

🌸 COMMANDS

• help — prints a usage summary 💜
• test — figures out the parameters that would be used to load the kdump kernel, but doesn't actually load it. Super useful for checking your config settings before you commit 🙈
• show — shows the current parameters and the saved kexec command from the last kdump-config load 👑
• status — checks /sys/kernel/kexec_crash_loaded or /sys/kernel/fadump_register (depending on the dump mode) and prints a matching message 🌈
• load — loads the kdump kernel, or registers it if you're using firmware assisted dump (fadump) 🔥
• unload — unloads the kdump kernel, or unregisters it if you're using firmware assisted dump (fadump) 🫶
• savecore — if /proc/vmcore exists, saves it using makedumpfile 🍓
• propagate — sends your public ssh key to a remote host so you can connect without a password 🦋
• symlinks {kernel-version} — verifies and creates the vmlinuz and initrd.img links in /var/lib/kdump for the kernel version you give it. If the target initrd.img file is missing, it creates that too ✨

📁 FILES

• /etc/init.d/kdump-tools — the init script that automatically loads a kdump kernel, or saves a vmcore and reboots 🔮
• /etc/default/kdump-tools — the configuration file 💅
• /var/crash/kernel_link — a link to the current debug kernel 🎀
• /var/crash/kexec_cmd — the last kexec_cmd that kdump-config ran 🌸
• /var/lib/kdump/vmlinuz — a symbolic link pointing to the current kernel file in /boot 💜
• /var/lib/kdump/initrd.img — a symbolic link pointing to the initrd.img file with the current kernel version in /boot 🧚

🚨 DIAGNOSTICS

Okay, these are the error messages you might run into, and what they actually mean. Don't panic, bestie 🙈

• USE_KDUMP is not set or zero — the variable USE_KDUMP in /etc/default/kdump-tools is 0. To use kdump, open the file and change the value to 1 💖
• kdump is not supported by this kernel — the file /sys/kernel/kexec_crash_loaded doesn't exist. This means the running kernel either wasn't configured to support kdump, or it was booted without a crashkernel= commandline parameter 💀
• no crashkernel= parameter in the kernel cmdline — a crashkernel= commandline parameter wasn't used when the running kernel booted. You'll need to add a crashkernel= parameter to your /etc/lilo.conf, /etc/elilo.conf, or grub /boot/grub/menu.lst file and reboot before kdump will work. Check the installed READMEs for suggestions 📚
• KDUMP_KERNEL does not exist: $KDUMP_KERNEL — a KDUMP_KERNEL was set in /etc/default/kdump-tools, but it can't be found 😭
• KDUMP_INITRD does not exist: $KDUMP_INITRD — a KDUMP_INITRD was set in /etc/default/kdump-tools, but it can't be found 😭
• no KDUMP_KERNEL in $KDUMP_DEFAULTS and can't find a kdump package directory in /usr/lib — no kdump kernel could be found anywhere 🫠
• found a linux-kdump package in $KDUMP_PKG_DIR but can't find a kdump kernel in $KDUMP_PKG_DIR — this one means the linux-kdump package is broken or installed wrong 🔧
• no kdump initrd image: $KDUMP_INITRD — an initrd will not be used: a kdump kernel was found, but its matching initrd wasn't 🥺
• kdump kernel relocation address does not match crashkernel= parameter — on non-relocatable architectures, the kdump kernel has to be built with a predetermined start address. This message means that start address doesn't match the one in the crashkernel= parameter 🙅‍♀️
• failed to load kdump kernel — kexec said it couldn't load the kdump kernel 💔
• failed to unload kdump kernel — kexec said it couldn't unload the kdump kernel. Run kexec -u directly for more details 🔍
• makedumpfile failed, falling back to 'cp' — the kdump-tools init script tried to save /proc/vmcore with makedumpfile, the save failed, and it's trying cp instead 🌸
• failed to save vmcore in $KDUMP_STAMPDIR — both makedumpfile and the cp fallback failed. This usually happens when pathnames are wrong or the disk is full 💀
• Invalid argument : missing kernel version — the symlinks command was used with a kernel version that isn't installed on this server 🙈
• Invalid symlink : {symbolic link} — the listed symbolic link is missing, broken, or pointing at the wrong version 🔗
• Cannot change symbolic links when kdump is loaded — the defined symbolic links are already in use. You have to unload kdump before using the symlinks command 🚫
• Unable to locate kernel hook — the kdump-tools script that creates the smaller initrd.img files is missing from /usr/kernel/postinst.d 🧩
• Invalid kernel version : {kernel version} — the kernel version you gave the symlinks command doesn't exist 🤷‍♀️
• fadump is not configured in this kernel. — Firmware Assisted Dump (fadump) isn't available on this platform 🥀
• SSH and NFS cannot be defined concurrently — both variables are set in /etc/default/kdump-tools. You have to pick one or the other 🙅‍♀️
• Unable to mount remote NFS directory {directory}. Cannot save core — the remote NFS server isn't letting this server mount the directory to save the crash dump 🚫
• Unable to write to the remote NFS directory {directory}. Cannot save core — the remote NFS directory mounts fine, but this server can't write files to it 😭
• makedumpfile --dump-dmesg failed. dmesg content will be unavailable — the command to capture the dmesg buffer failed, so it won't show up in the crash directory 🥺
• Unable to reach remote server {server}. No reason to continue — kdump-tools is set up for a remote dump, but the remote server can't be reached 📡
• makedumpfile scp failed. The vmcore file will not be available — makedumpfile worked, but the resulting file couldn't be copied over SSH to the remote server 💔
• Remote ssh dump is not configured. No reason to propagate" — you ran the propagate command, but kdump isn't set up for a remote dump 🙃

🔗 SEE ALSO

• /usr/share/doc/kdump-tools/README
• /usr/share/doc/kdump-tools/README.Debian
• kdump-tools(5), kexec(8), crash(8), gdb(1), makedumpfile(8) 🌈

👑 AUTHOR

Written by Terry Loftin <terry.loftin@hp.com> and Louis Bouchard <louis.bouchard@ubuntu.com> 🫶

kdump-tools v1.6.7-2 · Jul 2020 · kdump-config(8) 🌸✨💅 bestie, you're officially a kdump expert now 🦋