Instance Monitoring and Health

Overview

Every ClawHosters instance includes built-in health monitoring. The instance dashboard shows the current status of your instance, resource usage, and container health. This page explains how health checks work, what the status indicators mean, and what to do when something goes wrong.

Instance Status

The instance dashboard displays the current status of your instance with a color-coded indicator:

A running instance with a green indicator means the container is up and responding to health checks.

Docker Health Checks

Each instance runs inside a Docker container with an automatic health check configured:

Docker marks the container as healthy when the check command succeeds, or unhealthy after 3 consecutive failures. The 60-second start period gives the OpenClaw gateway time to initialize before health checks begin.

Resource Usage

The instance dashboard displays resource usage for your VPS:

These metrics update periodically. If any metric consistently reaches high levels, your workload may benefit from a higher tier.

Resource Limits by Tier

The memory limit is enforced at the Docker container level. If your instance exceeds this limit, Docker may terminate the container process (OOM kill).

Checking Health via SSH

If you have SSH access enabled, you can inspect your instance directly:

# Check if the container is running
docker ps

# View container health status
docker inspect --format='{{.State.Health.Status}}' openclaw-<instance_id>

# View recent health check results
docker inspect --format='{{json .State.Health}}' openclaw-<instance_id> | python3 -m json.tool

# Check resource usage
docker stats --no-stream

# Check disk space on the VPS
df -h

# Check memory usage
free -h

Replace <instance_id> with your instance ID shown in the dashboard.

Common Health Issues

Container Restart Loops

The container keeps stopping and restarting. Possible causes:

• The OpenClaw process crashes on startup due to a configuration issue
• Insufficient memory causes an OOM kill, and Docker restarts the container

Check the container logs for error messages:

docker logs openclaw-<instance_id> --tail 50

Out-of-Memory (OOM) Kills

Docker terminates the container when it exceeds its memory limit. Signs of OOM:

• Container restarts unexpectedly
• docker inspect shows OOMKilled: true

If this happens frequently, consider upgrading to a higher tier with more memory.

Disk Full

When the VPS runs out of disk space, the container may fail to write logs or data. Check with:

df -h

If disk usage is near capacity, check for large log files or accumulated data. Docker log rotation is configured (50 MB max, 3 files), but application data inside the container can still grow.

Health Check Failures

The health check calls http://localhost:18789/ inside the container. If this fails:

• The OpenClaw gateway may still be starting (wait 60 seconds after a restart)
• The gateway process may have crashed (check container logs)
• Port binding issues inside the container

What to Do When Your Instance is Unhealthy

Step 1: Check the Dashboard

Open your instance dashboard. The status indicator and any error messages give the first clue.

Step 2: Restart the Instance

Most transient issues are resolved by restarting:

1. Go to the Instance Dashboard
2. Click Stop
3. Wait a few seconds
4. Click Start

The instance reprovisions and the container starts fresh.

Step 3: Rebuild the Instance

If restarting does not help, rebuild the instance from the dashboard. This creates a fresh container with your existing configuration. Custom modifications made via SSH will be lost.

Step 4: Check Logs (SSH)

If you have SSH access, inspect the container logs:

docker logs openclaw-<instance_id> --tail 100

Look for error messages, stack traces, or repeated crash patterns.

Step 5: Contact Support

If the issue persists and you cannot identify the cause, contact support through the ClawHosters dashboard. Include:

• Your instance ID
• The status shown in the dashboard
• When the issue started
• Any recent changes you made

Support is available for instances without SSH enabled. Instances with SSH access are limited to billing and account support.

Related Documentation

• Instance Overview — Instance statuses, lifecycle, and tiers
• Instance Settings and Configuration — LLM, messaging, and web access settings
• Starting and Stopping Instances — Power management and billing impact
• SSH Access to Your Instance — Direct server access for advanced users
• Troubleshooting Common Issues — Solutions for frequent problems