A GoLang based HTTP server which will send a Wake-on-LAN package (magic packet) on a local network. The request can be send using web interface or directly using HTTP request with the mapped device name in the URL. The only computing device I have running 24x7 is handy-dandy Raspberry Pi 4 (4GB) with docker containers. All other devices like server, laptop and NAS as powered only when I need them. I needed a way to easily turn them on specifically when trying to automate things like nightly builds.

This application is intended do be used in conjunction with a reverse proxy and secured with an SSL certificate. As the intended use case was with home networks, the application has no in-built authentication. While this could pose a slight security risk even if this was hacked to application is intended to be containerised so the attack surface if limited.

I have bookmarked direct link to device(s) on my browsers to wake them using single HTTP call for ease of access.

Use cases:

• Wake-up my home computers remotely, for access remotely over RDP.
• Integration with automated routines to allow parts of a home lab to sleep instead of running 24x7 to save energy.

On some devices WOL can be difficult to correctly configure and have work reliably.

Follow this article for Dell Laptops.

The UI features CRUD operation implemented using js-grid.com plugin.

/wolweb/wake/<hostname> - Returns a JSON object

{
  "success": true,
  "message": "Sent magic packet to device Server with Mac 34:E6:D7:33:12:71 on Broadcast IP 192.168.1.255:9",
  "error": null
}

The application will use the following default values if they are not explicitly configured as explained in sections below.

You can override the default application configuration by using a config file or by setting environment variables. The application will first load values from config file and look for environment variables and overwrites values from the file with the values which were found in the environment.

Using config.json:

{
    "host": "0.0.0.0",
    "port": 8089,
    "vdir":"/wolweb",
    "bcastip":"192.168.1.255:9"
}

Using Environment Variables:

Environment variables takes precedence over values in config.json file.

{
    "devices": [
        {
            "name": "Server",
            "mac": "34:E6:D7:33:12:71",
            "ip": "192.168.1.255:9"
        },
        {
            "name": "NAS",
            "mac": "28:C6:8E:36:DC:38",
            "ip": "192.168.1.255:9"
        },
        {
            "name": "Laptop",
            "mac": "18:1D:EA:70:A0:21",
            "ip": "192.168.1.255:9"
        }
    ]
}

This project includes Dockerfile (based on Alpine) and docker-compose.yml files which you can use to build the image for your platform and run it using the docker compose file. If interested, I also have alternate Dockerfile (based on Debian). Both of these Dockerfile are tested to run on Raspberry Pi Docker CE. If you want to use this application as-is, you will only need to download these two docker-related files to get started. The docker file will grab the code and compile it for your platform.

I could not get this to run using Docker's bridged network. The only way I was able to make it work was to use host network for the docker container. See this docker/for-linux#637 for details.

docker-compose up -d

docker build -t wolweb .
docker run --network host -it wolweb

docker cp wolweb:/wolweb - > wolweb.gz

You need Go 1.20 to build binary for any OS.

# Windows
go build -o wolweb.exe .

# Linux/macOS
go build -o wolweb .

I initially thought of running this application on my router, so I needed to build the application without having to install build tool on my router. I use the following PowerShell one liner to build targeting the ARM v5 platform on my Windows machine with VS Code:

 $Env:GOOS = "linux"; $Env:GOARCH = "arm"; $Env:GOARM = "5"; go build -o wolweb .

Copy the file over to router and make it executable.

chmod +x wolweb

To see detailed instructions on how to run this application as service on ASUS router with custom firmware asuswrt-merlin see this Wiki guide

I am already using NGiNX as web-proxy for accessing multiple services (web interfaces) from single IP and port 443 using free Let's Encrypt HTTPS certificate. For accessing this service, I just added the following configuration under my existing server node.

location /wolweb {
	proxy_pass http://192.168.1.4:8089/wolweb;
}

This is also the reason why I have an option in this application to use virtual directory /wolweb as I can easily map all requests for this application. My / is already occupied for other web application in my network.

This project is based on a couple of framworks and provided below:

• https://github.com/dabondi/go-rest-wol - Provided the initial project framework
• https://github.com/sabhiram/go-wol - Provided the functionality to the REST

Thank you to the developers behind both projects, David Baumann and Shaba Abhiram!

Distributed with GNU General Public License (c) 2023