How to Run OpenClaw 24/7

By Clawculus · Last updated February 2026 · 10 min read

Your AI agent isn't very useful if it only works when your laptop is open. This guide covers everything you need to run OpenClaw continuously — from VPS selection to auto-restart to heartbeat configuration.

Why Run 24/7?

An always-on agent can:

• Respond to messages on WhatsApp, Telegram, Discord any time
• Run heartbeat checks (email, calendar, notifications)
• Execute scheduled cron tasks
• Maintain long-running background processes
• Accept webhooks and API calls from other services

Option 1: VPS with Docker (Recommended)

The gold standard for 24/7 operation. A VPS (Virtual Private Server) gives you a dedicated Linux box in the cloud that never sleeps.

Recommended Providers

Quick Start with Docker

# SSH into your VPS
ssh root@YOUR_VPS_IP

# Install Docker
curl -fsSL https://get.docker.com | sh

# Clone OpenClaw
git clone https://github.com/openclaw/openclaw.git
cd openclaw

# Create persistent directories
mkdir -p /root/.openclaw/workspace
chown -R 1000:1000 /root/.openclaw

# Run the setup script
./docker-setup.sh

The docker-setup.sh script handles everything: builds the image, runs onboarding, generates a gateway token, and starts the gateway. Docker Compose's restart: unless-stopped policy means it survives reboots automatically.

Verify It's Running

# Check container status
docker compose ps

# View logs
docker compose logs -f openclaw-gateway

# Health check
docker compose exec openclaw-gateway node dist/index.js health --token "$OPENCLAW_GATEWAY_TOKEN"

Option 2: Bare Metal with systemd

If you prefer running directly on the host without Docker, use systemd to manage the process.

Install OpenClaw

# Install Node.js 22+
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -
sudo apt-get install -y nodejs

# Install OpenClaw globally
npm install -g openclaw

# Run onboarding
openclaw onboard

Create a systemd Service

sudo cat > /etc/systemd/system/openclaw.service << 'EOF'
[Unit]
Description=OpenClaw Gateway
After=network.target

[Service]
Type=simple
User=node
WorkingDirectory=/home/node
ExecStart=/usr/bin/openclaw gateway
Restart=always
RestartSec=10
Environment=NODE_ENV=production

[Install]
WantedBy=multi-user.target
EOF

sudo systemctl daemon-reload
sudo systemctl enable openclaw
sudo systemctl start openclaw

Manage the Service

sudo systemctl status openclaw    # Check status
sudo systemctl restart openclaw   # Restart
sudo journalctl -u openclaw -f   # View logs

Option 3: Screen/tmux (Quick & Dirty)

For testing or temporary setups, just run in a detached terminal session:

# Start a detached screen session
screen -dmS openclaw openclaw gateway

# Reattach to check on it
screen -r openclaw

Setting Up Heartbeats

Once running 24/7, configure heartbeats so your agent proactively checks in:

// In openclaw.json
{
  "agents": {
    "defaults": {
      "heartbeat": {
        "every": "30m",
        "target": "last",
        "activeHours": {
          "start": "08:00",
          "end": "23:00"
        }
      }
    }
  }
}

Key settings:

• every: How often to poll (e.g. "20m", "1h")
• target: Where to send alerts — "last" uses the last channel you chatted on, "none" runs silently
• activeHours: Don't wake you up at 3am (optional)

Create a HEARTBEAT.md in the workspace to tell the agent what to check:

# Heartbeat Checklist
- Check for unread emails
- Review upcoming calendar events (next 24h)
- If anything is urgent, send an alert
- Otherwise, reply HEARTBEAT_OK

Accessing Your Agent Remotely

Once deployed, access the Control UI via SSH tunnel:

# From your laptop
ssh -N -L 18789:127.0.0.1:18789 root@YOUR_VPS_IP

# Then open in browser
# http://127.0.0.1:18789/

Or connect your messaging channels (WhatsApp, Telegram, Discord) and chat with your agent from anywhere.

Monitoring & Staying Alive

• Docker: restart: unless-stopped handles crashes and reboots
• systemd: Restart=always with RestartSec=10
• Health endpoint: Use openclaw health from a separate cron to verify
• Log rotation: Docker handles this; for systemd, configure journald limits
• Disk space: Watch for chat history and media growth over time