Skip to content

Troubleshooting and Additional Deployment Tips

Diagnosing Docker Issues

Check Docker Service state:

sudo systemctl status docker
  • Inactive/failed → service issue.
  • Active but CLI fails → daemon issue.

Service Commands:

sudo systemctl start docker      # Starts the service and daemon
sudo systemctl stop docker       # Stops the service and daemon
sudo systemctl status docker     # Shows service + daemon status
sudo systemctl enable docker     # Auto-start daemon at boot

Restart Docker Daemon:

sudo systemctl restart docker
pgrep -x dockerd || sudo dockerd

If dockerd is run manually, it stops when terminal closes. Enable it for auto-start.

Run MCP Server in Background

docker run -d --name <your-container-name> \
  -e H3_API_KEY={your-key-here} \
  horizon3ai/h3-mcp-server:latest

Stop later:

docker stop <your-container-name>

Clean Up Containers and Keys

docker rm <your-container-name>
unset H3_API_KEY

Use Single-User Mode

Each container should use one API key.

Available Tools

The MCP server provides the following tools to AI assistants:

  1. fetch_h3_graphql_docs: Get GraphQL schema documentation for constructing queries
  2. run_h3_graphql_request: Execute GraphQL queries against the H3 API