Stop Paying Full Reboot Downtime: Practical systemd soft-reboot on Linux

Description and use

Using systemd soft-reboot restarts only userspace while keeping the kernel and hardware state, reducing downtime after package updates.

Problem

After a userspace package update, the system remains running old libraries while the kernel stays unchanged, causing unnecessary full reboot downtime.

Outcome

Using systemd soft-reboot reduces downtime by restarting only userspace while keeping the kernel and hardware state intact, enabling quick recovery after package updates and optional root flips.

Workflow

• 1. Verify systemd version ≥254 and presence of soft-reboot.target.
• 2. Ensure root or polkit rights to run power commands.
• 3. (Optional) Prepare /run/nextroot/ with a new root filesystem if a root flip is needed.
• 4. Run sudo systemctl soft-reboot (or with --when/--force as needed).
• 5. Systemd isolates to soft-reboot.target, stops services, unmounts shutdown graph mounts, then reexecutes PID 1.
• 6. If /run/nextroot/ exists, kernel switches root before reexec.
• 7. New userspace boot transaction starts; services restart under the same kernel.
• 8. Verify by checking kernel release (uname -r unchanged) and timestamps (UserspaceTimestamp advanced).

Evidence

systemctl soft-reboot was added in systemd 254.

Source

soft-reboot isolates toward soft-reboot.target, which requires systemd-soft-reboot.service.

Source

If /run/nextroot/ exists, the root filesystem is switched to it.

Source