OpenClaw systemd Service: Auto-Start, Crash Recovery & Health Monitoring (2026)

OpenClaw systemd Service: Auto-Start, Crash Recovery & Health Monitoring

You’ve deployed OpenClaw on a VPS — great. But what happens when the process crashes at 3 AM? Or when the server reboots after a kernel update? If you’re still running OpenClaw in a tmux session, this guide is for you.

We’ll set up OpenClaw as a proper systemd service with automatic restart, boot-time startup, health monitoring, and log management. Total setup time: ~10 minutes.

Prerequisites

• A Linux VPS (Ubuntu 22.04/24.04, Debian 12, or similar) — Vultr and DigitalOcean both work great for this
• OpenClaw already installed and working manually
• Root or sudo access

Step 1: Create the systemd Service File

sudo nano /etc/systemd/system/openclaw.service

Paste this config (adjust User and paths to match your setup):

[Unit]
Description=OpenClaw AI Agent Gateway
After=network-online.target
Wants=network-online.target

[Service]
Type=simple
User=your-username
Group=your-username
WorkingDirectory=/home/your-username
ExecStart=/usr/local/bin/openclaw gateway start --foreground
Restart=always
RestartSec=5
StartLimitIntervalSec=300
StartLimitBurst=10

# Environment
Environment=NODE_ENV=production
Environment=HOME=/home/your-username

# Security hardening
NoNewPrivileges=true
ProtectSystem=strict
ReadWritePaths=/home/your-username

# Resource limits
MemoryMax=2G
TasksMax=100

# Logging
StandardOutput=journal
StandardError=journal
SyslogIdentifier=openclaw

[Install]
WantedBy=multi-user.target

Key Settings Explained

Step 2: Enable and Start

# Reload systemd to pick up the new service
sudo systemctl daemon-reload

# Enable auto-start on boot
sudo systemctl enable openclaw

# Start the service now
sudo systemctl start openclaw

# Check it's running
sudo systemctl status openclaw

You should see Active: active (running). If not, check the logs:

journalctl -u openclaw -n 50 --no-pager

Step 3: Common Management Commands

# Stop the service
sudo systemctl stop openclaw

# Restart (e.g., after config changes)
sudo systemctl restart openclaw

# View live logs (like tail -f)
journalctl -u openclaw -f

# View logs from last hour
journalctl -u openclaw --since "1 hour ago"

# View logs from last boot
journalctl -u openclaw -b

# Check if auto-start is enabled
systemctl is-enabled openclaw

Step 4: Health Check Script

Create a simple health check that alerts you if OpenClaw is down:

nano ~/openclaw-health-check.sh

#!/bin/bash
# OpenClaw Health Check
# Add to crontab: */5 * * * * /home/your-username/openclaw-health-check.sh

SERVICE="openclaw"
LOG_FILE="/home/your-username/openclaw-health.log"

if ! systemctl is-active --quiet "$SERVICE"; then
    TIMESTAMP=$(date '+%Y-%m-%d %H:%M:%S')
    echo "[$TIMESTAMP] OpenClaw is DOWN — systemd should auto-restart" >> "$LOG_FILE"
    
    # Optional: send alert via curl to your webhook
    # curl -s -X POST "https://your-webhook-url" \
    #   -H "Content-Type: application/json" \
    #   -d "{\"text\": \"⚠️ OpenClaw service is down on $(hostname). Auto-restart triggered.\"}"
fi

chmod +x ~/openclaw-health-check.sh

# Add to crontab (runs every 5 minutes)
(crontab -l 2>/dev/null; echo "*/5 * * * * /home/your-username/openclaw-health-check.sh") | crontab -

Step 5: Log Rotation

Prevent logs from eating all your disk space:

sudo nano /etc/logrotate.d/openclaw

/var/log/journal/*openclaw* {
    daily
    rotate 7
    compress
    missingok
    notifempty
}

Or limit journal size globally:

sudo journalctl --vacuum-size=500M

To make the limit permanent:

sudo nano /etc/systemd/journald.conf
# Add or uncomment:
# SystemMaxUse=500M
sudo systemctl restart systemd-journald

Step 6: Automatic Updates (Optional)

If you want OpenClaw to auto-update on a schedule:

nano ~/openclaw-update.sh

#!/bin/bash
# Weekly OpenClaw update
LOG="/home/your-username/openclaw-update.log"
TIMESTAMP=$(date '+%Y-%m-%d %H:%M:%S')

echo "[$TIMESTAMP] Starting update..." >> "$LOG"

# Update OpenClaw
npm update -g openclaw 2>&1 >> "$LOG"

# Restart service to pick up new version
sudo systemctl restart openclaw

echo "[$TIMESTAMP] Update complete" >> "$LOG"

chmod +x ~/openclaw-update.sh

# Run weekly on Sunday at 4 AM
(crontab -l 2>/dev/null; echo "0 4 * * 0 /home/your-username/openclaw-update.sh") | crontab -

Troubleshooting

Service fails to start immediately

# Check the exact error
journalctl -u openclaw -n 30 --no-pager

# Common issues:
# - Wrong User/Group in service file
# - Wrong path in ExecStart
# - Port already in use (another OpenClaw instance running)

If the port is in use:

# Find what's using the port
sudo lsof -i :3000
# Kill it
sudo kill -9 <PID>
# Then restart
sudo systemctl restart openclaw

Service keeps restarting (crash loop)

# Check restart count
systemctl show openclaw --property=NRestarts

# If hitting StartLimitBurst, reset the counter
sudo systemctl reset-failed openclaw
sudo systemctl start openclaw

Service started but gateway not responding

# Verify the gateway is actually listening
openclaw gateway status

# Check if the foreground flag is needed for your version
openclaw gateway start --help

Permission errors

# Ensure the user owns the workspace
sudo chown -R your-username:your-username /home/your-username/.openclaw

Recommended VPS Providers

For a reliable 24/7 OpenClaw deployment, you need a VPS with good uptime and network connectivity:

• Vultr — From $6/mo, fast SSD, 17 global locations. Great for latency-sensitive setups.
• DigitalOcean — From $6/mo, excellent monitoring dashboard, easy snapshots for backup.
• Tencent Cloud — Best choice if you’re in mainland China or need low-latency Asia coverage. Competitive pricing on lightweight instances.

A 1 vCPU / 1 GB RAM instance handles a single-user OpenClaw deployment comfortably. Upgrade to 2 GB if you plan to run multiple agents or heavy browser automation.

Summary

Your OpenClaw agent now survives crashes, reboots, and updates automatically. No more babysitting tmux sessions.

Need help deploying OpenClaw from scratch? Check our VPS Deployment Guide for the full walkthrough.

Related Reading

• Silent Message Loss / Replay Troubleshooting (2026)