Instance Won't Start

When Your Instance Refuses to Start

If your instance is stuck or fails to start after pressing "Start" in the dashboard, work through the sections below in order. Most startup failures have straightforward causes.

Check Your Balance

The most common reason an instance won't start is insufficient balance. The system requires enough Claws to cover at least one day of operation before starting an instance.

What to check:

• Open your dashboard and look at your Claws balance
• Compare it against the daily rate for your tier (visible on the instance page)
• If your balance is below the daily rate, add funds first

If your instance was paused due to zero balance, it will resume from a snapshot once you add enough Claws. This can take 30-60 seconds as the snapshot is restored to a new VPS.

Check Instance Status

The dashboard shows the current status of your instance. Here is what each status means when your instance is not running:

Status	What It Means	What to Do
Paused	Stopped due to low balance, snapshot saved	Add funds, then click "Resume"
Deploy Failed	The deployment step failed after provisioning	Click "Retry" on the instance page
Provisioning	VPS is still being created	Wait 1-2 minutes
Deploying	Software is being installed on the VPS	Wait 30-60 seconds
Suspended	Admin has suspended the instance	Contact support

If you see "Deploy Failed," the system already tried up to two automatic retries. Clicking "Retry" triggers a fresh deployment attempt.

Snapshot Resume Failures

If your instance was paused and you are trying to resume it, the system restores from a Hetzner snapshot. This can fail if:

• The snapshot was created from a corrupted state
• The snapshot expired (snapshots are kept for 7 days after pause, then deleted)

If the snapshot is expired: Your instance data is gone. You will need to create a new instance. This is why keeping your balance above zero is important.

If the snapshot restore fails: Click "Retry" on the instance page. If it fails again, contact support with your instance name.

Instance Starts But Shows Unhealthy

If the instance moves to "Running" but the health check shows issues:

• The container may be starting up. The OpenClaw gateway takes up to 60 seconds to become responsive.
• Wait one minute and refresh the page.
• If it stays unhealthy after two minutes, try rebuilding the instance from the dashboard.

Instance Was Working, Now Won't Start

If your instance was running fine and suddenly stopped:

1. Check the dashboard for status changes -- It may have been paused due to balance depletion
2. Check for Hetzner maintenance -- Rare, but Hetzner sometimes reboots VPS for host maintenance
3. Try restarting from the dashboard -- Use the "Restart" button
4. Try rebuilding -- This creates a fresh container but keeps your data volume

When Nothing Works

If you have tried all the steps above and your instance still won't start:

1. Note your instance name and the current status shown on the dashboard
2. Open a support ticket through the dashboard
3. Include what you have tried and any error messages you see

Related Docs

• Troubleshooting Common Issues -- Provisioning, deployment, and gateway problems
• Instance Overview -- How instances work and their lifecycle
• Start, Stop, and Restart -- Instance power management