When examining Linux system reboots at the kernel level, we encounter two distinct approaches for immediate system restarts that bypass normal shutdown procedures:

# Method 1: SysRq trigger
echo b > /proc/sysrq-trigger

# Method 2: Reboot command
reboot -n

The sysrq-trigger approach directly invokes kernel panic handling routines. This is what happens internally:

1. Immediately triggers a CPU reset through arch-specific low-level code
2. Completely bypasses all filesystem synchronization
3. No attempt to communicate with init system

In contrast, reboot -n operates through the normal syscall path but with modifications:

SYSCALL_DEFINE4(reboot, int, magic1, int, magic2, unsigned int, cmd, void __user *, arg)
{
    if (!(cmd & LINUX_REBOOT_CMD_NO_SYNC))
        kernel_power_off();

Consider these examples where the difference matters:

Filesystem Corruption Risk:
On an ext4 filesystem with journaling:

$ strace reboot -n
...
sync()                              # Still occurs at driver level
reboot(LINUX_REBOOT_CMD_RESTART)    # Final syscall

Versus sysrq-trigger which shows zero sync activity in kernel logs.

Use echo b > /proc/sysrq-trigger when:

• Debugging kernel panics
• Testing crash recovery systems
• Emergency hardware resets

Use reboot -n for:

• Fast restarts during development
• When some driver-level syncing is acceptable
• Scripted reboots needing partial cleanup

Behavior has evolved across kernel versions:

For emergency reboot scripts:

#!/bin/bash
# Prefer sysrq for true emergency cases
if [ "$EMERGENCY" = "true" ]; then
    echo 1 > /proc/sys/kernel/sysrq
    echo b > /proc/sysrq-trigger
else
    reboot -n
fi

When you need to force an immediate reboot on Linux systems (particularly RHEL-based distros like CentOS and Fedora), two common commands emerge:

# Method 1: Using reboot command
sudo reboot -n

# Method 2: Using sysrq trigger
echo b > /proc/sysrq-trigger

Both methods skip the normal filesystem sync operations, but with crucial differences:

• reboot -n: The -n flag tells the system not to explicitly call sync(), but the kernel and storage drivers may still perform synchronization.
• sysrq trigger: The 'b' command forces an immediate hardware-level reboot, bypassing all higher-level shutdown procedures.

The sysrq method operates at a lower level in the kernel's emergency management system:

# Check if sysrq is enabled (should return 1)
cat /proc/sys/kernel/sysrq

# Enable all sysrq functions temporarily
echo 1 > /proc/sys/kernel/sysrq

When to use reboot -n:

• When you need a quick reboot but want to maintain some system integrity
• For controlled testing environments

When to use sysrq trigger:

• During kernel panics when normal reboot fails
• For hardware testing that requires complete bypass of software layers
• Emergency situations where system is completely unresponsive

Both methods can lead to filesystem corruption. Always consider these alternatives first:

# Fast but safer reboot
sudo systemctl reboot --force --no-sync

# Emergency reboot with logging
echo s > /proc/sysrq-trigger  # sync disks
echo u > /proc/sysrq-trigger  # remount read-only
echo b > /proc/sysrq-trigger  # reboot

Modern systems using systemd handle these commands differently:

# View systemd's reboot behavior
systemctl show reboot.target

# Alternative systemd emergency reboot
systemctl --force --force reboot