Increase Docker Shutdown Timeout in Systemd

[szaimen]

When a Linux host reboots or the Docker service stops, systemd manages the shutdown sequence. Even if Docker is configured to let containers shut down gracefully, systemd will forcibly kill the entire Docker daemon if it takes longer than the systemd timeout limit.
Follow this guide to give systemd enough time to let Docker shut down your containers safely.

Prerequisites

• Access to a Linux terminal.
• sudo or root privileges on the host machine.

---

Step 1: Create a Systemd Service Override

Instead of editing the core Docker service file (which updates can overwrite), use systemd's built-in override mechanism.
Run the following command to open the override configuration file in your default text editor:

sudo systemctl edit docker.service

Step 2: Add the Timeout Configuration

A blank file will open (or a file with commented-out template text). Insert the following configuration block at the very top of the file:

[Service]
TimeoutStopSec=30m

💡 Tip: Change 30m to whatever your longest-running container needs to safely stop. You can also use time units like 2m for two minutes or 90s for 90 seconds.

Save and close the text editor:

• If using Nano: Press Ctrl + O then Enter to save, and Ctrl + X to exit.
• If using Vim: Type :wq and press Enter.

Step 3: Reload Systemd and Apply Changes

Systemd needs to register the new configuration change. Run the daemon-reload command, then restart the Docker service to apply the new timeout window:

# Reload the systemd manager configuration
sudo systemctl daemon-reload
# Restart Docker to apply the changes
sudo systemctl restart docker

Step 4: Verify the New Configuration

To ensure systemd successfully accepted your new timeout setting, check the runtime attributes of the Docker service:

systemctl show docker.service | grep TimeoutStopSec

Expected Output:

TimeoutStopSecUSec=30min

(Note: Systemd often displays the verified time in microseconds or human-readable formats like 2min).

---

Troubleshooting & Verification

If your host reboots still result in hard-killed containers, you can verify how long Docker is taking to stop by inspecting the system journal after a restart:

journalctl -u docker.service -b -1

(This command shows the Docker service logs from the previous boot cycle to help identify which container is causing the delay).