mcp-server-uyuni-report

jordimassaguerpla/mcp-server-uyuni-report

3.1

If you are the rightful owner of mcp-server-uyuni-report and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The MCP Server for Uyuni is a protocol server designed to facilitate communication and management tasks within the Uyuni systems management platform.

MCP Uyuni Report Server

This MCP server provides tools to query and analyze data from a Uyuni report database, offering insights into patch management and system compliance.

Disclaimer

This is an open-source project provided "AS IS" without any warranty, express or implied. Use at your own risk. For full details, please refer to the License section.

Available Tools

  • get_errata_stats
  • get_filter_values
  • plot_errata_stats_timeseries

get_errata_stats

This is the main tool for retrieving statistics about errata. It can provide either a single aggregate summary or a time-series of data suitable for plotting.

Example 1: Getting an Aggregate Summary

"What are the errata statistics for all 'critical' security advisories in the 'Production Servers' organization during the first quarter of this year?"

Example 2: Generating a Time-Series Graph

"Show me a weekly trend of security errata for my 'sles' systems from the beginning of the year until now. Plot the mean patch time, patched systems, and created errata on a graph."

Filter Examples

You can use various filters to narrow down your query. Here are examples for each filter individually and how to combine them.

Individual Filters
  • Date Range:
    • start_date: "What are the errata statistics for patches issued since the start of 2024?"
    • end_date: "Show me errata stats for patches issued before July 1st, 2023."
  • Advisory Properties:
    • advisory_type: "Get me the stats for all 'bugfix' advisories."
    • advisory_name: "What are the stats for advisories with 'kernel' in their name?"
    • cve: "Find statistics for the errata related to 'CVE-2023-12345'."
    • severity: "What are the statistics for all 'moderate' severity errata?"
  • System Properties:
    • hostname: "Show me errata statistics for systems with 'db-server' in their hostname."
    • os_family: "Get statistics for all 'rhel' systems."
    • organization: "What are the errata stats for the 'Human Resources' organization?"
    • system_group_name: "Show me the statistics for the 'PCI-Compliance' system group."
  • Output Format:
    • granularity: "Show me a monthly trend of all errata from the beginning of the year until now."
Combining Filters

Filters can be combined to create highly specific queries.

  • Aggregate Example:

    "What are the aggregate statistics for 'critical' security patches related to CVEs containing '2024' that affect 'sles' systems in the 'Finance' organization?"

  • Time-Series Example:

    "Show me a weekly trend of 'important' security errata for systems in the 'Web Servers' group, from the beginning of the year until now."

  • Complex Aggregate Example:

    "Calculate the mean time to resolution for 'bugfix' advisories with 'samba' in their name, applied to systems in the 'File Servers' group within the 'Engineering' organization, for patches issued in the last 6 months."

get_filter_values

This helper tool allows you to discover the possible values for the various filters in the get_errata_stats tool.

Example 1: Discovering Available Organizations

"What organizations can I filter by?"

Example 2: Searching for a Specific Hostname

"Show me available hostnames that contain 'web-server'."

plot_errata_stats_timeseries

This tool directly generates a PNG image of the time-series data, which can be displayed by a capable client.

Example

"Plot a monthly graph of all security errata for the 'Finance' organization over the last year."

Running with an MCP Client

To use this server with an MCP client (such as mcp-host, gemini-cli, openwebui, etc.), you need to configure the client to know how to start the mcp-server-uyuni-report process. This is done by creating a config.json file in your client's configuration directory.

This server requires its own configuration to connect to the database. This can be provided via a separate configuration file (e.g., .env or config) containing environment variables.

Configuration

The server is configured using environment variables. These can be placed in a file (e.g., config or .env) that is then passed to the uv or docker command.

Required Environment Variables
  • DB_USER: The username for the database connection.
  • DB_PASSWORD: The password for the database connection.
Optional Environment Variables
  • DB_HOSTNAME: The database server hostname (default: localhost).
  • DB_PORT: The database server port (default: 5432).
  • DB_NAME: The database name (default: reportdb).
  • LOG_FILE: Path to a file where logs will be written (e.g., /var/log/mcp-uyuni-report.log). If not set, logs are written to standard error.
  • LOG_LEVEL: The logging level for the application (e.g., DEBUG, INFO, WARNING). Default: INFO.

Method 1: Using uv (Local Development)

This method runs the server directly on your machine using uv.

  1. Create a server configuration file with the necessary environment variables for the database connection. For example, create a file named config in the project root:

    # File: /path/to/mcp-server-uyuni-report/config
DB_USER="your_db_user"
DB_PASSWORD="your_db_password"
DB_HOSTNAME="localhost"
DB_NAME="reportdb"
LOG_LEVEL="DEBUG"

  2. Create a client config.json file. This file tells your MCP client how to launch the server using uv and where to find the server's configuration file.

    Example config.json for your client:

    {
  "mcpServers": {
    "mcp-server-uyuni-report": {
      "command": "uv",
      "args": [
        "run",
        "--env-file=/path/to/mcp-server-uyuni-report/config",
        "--directory",
        "/path/to/mcp-server-uyuni-report",
        "mcp-server-uyuni-report"
      ]
    }
  }
}
    • Replace /path/to/mcp-server-uyuni-report with the absolute path to this project's directory.

Method 2: Using a Container (Docker/Podman)

This method runs the server inside a container, which is ideal for production or isolated environments. The setup is similar, but the client's config.json will use docker or podman to run the server.

  1. Create the server configuration file as described in the uv method above.

  2. Create a client config.json file that instructs the client to launch a container.

    Example config.json for your client:

    {
  "mcpServers": {
    "mcp-server-uyuni-report": {
      "command": "docker",
      "args": [
        "run", "--rm", "-i",
        "--env-file", "/path/to/mcp-server-uyuni-report/config",
        "mcp-uyuni-report-server:latest"
      ]
    }
  }
}
    • This assumes you have built a Docker image named mcp-uyuni-report-server:latest.
    • Ensure the path to the --env-file is correct and accessible from where the client is run.

Using a local database

The recommended workflow is to start a local database and import the data from a reportdb dump. This way, the MCP Server does not need direct access to the reportdb database.

Staring the database

You can start the postgresql server:

./start_uyuni_report_db.sh

Importing data

In your uyuni server run:

podman exec -ti uyuni-server pg_dump --username USERNAME --dbname=reportdb --host=localhost --port=5432 | gzip > test_data.sql.gz

Then, copy the test_data.sql.gz into your running directory and

gunzip test_data.sql.gz

And load the data

./insert_test_data.sh test_data.sql

Feedback

We would love to hear from you! Any idea you want to discuss or share, please do so at [https://github.com/uyuni-project/uyuni/discussions/10562](https://github.com/uyuni-project/uyuni/disc ussions/10562)

If you encounter any bug, be so kind to open a new bug report at [https://github.com/uyuni-project/mcp-server-uyuni/issues/new?type=bug](https://github.com/uyuni-project/mcp-server-uyuni/is sues/new?type=bug)

Thanks in advance from the uyuni team!

License

This project is licensed under the Apache License, Version 2.0. See the file for details.