WOL Proxy

The "WOL Proxy" is intended to be used as a proxy for a website, that displays fallback web service when the server that runs the actual website is sleeping or shut down. This service can then be used to up the server via WakeOnLan.

Our use case is a server, that is rarely accessed, but needs lots of power even when idling. The powerful server has a script, that stops the server if it hasn't been accessed for a certain period of time. Nginx is running as reverse-proxy on another server on the same network configured similar to the nginx config below. It serves the webpage of the powerful server as long as it's accessible and uses a locally running WOL Proxy instance as backup server. When nginx gets a request and the powerful server isn't running it automatically switches to the backup server and displays the start button of the WOL Proxy. After the powerful server is started nginx detects after around 10 seconds that it is available again and switches back to the original web page.

Config

The config has the following parameters:

config key	env variable name	example	explanation
listenAddress	LISTEN_ADDRESS	0.0.0.0:8080	The address the service listens on.
broadcastAddress	BCAST_ADDRESS	192.168.0.255	The address the WOL packet is broadcasted. Should be the broadcast address of your subnet.
macAddress	MAC_ADDRESS	01:23:45:67:89:AB	The MAC address of the target computer (needed for the magic WOL packet)
serviceName	SERVICE_NAME	My Service	The name of your service. Will be displayed on the webpage.
serverWakeupTime	SERVER_WAKEUP_TIME	60	The time in seconds the server needs to start + approx. 10 seconds for nginx to register that the service is working again.
serviceAliveRoute	SERVICE_ALIVE_ROUTE	/api/v1/status	A route on which the original service returns a status code that is different to 425 TooEarly. The client uses it to automatically reload the webpage.

see config.example.yml in the repo

nginx Config

Here is an example nginx config with explanations:

http {
    # define a new backend
    upstream failover_backend {
        # this is the default address
        server othermachine:8080;
        # this is the backup address that is only used if the primary doesn't work
        server localhost:8080 backup;
    }

    server {
        listen       80;
        server_name  localhost;

        location / {
            # reverse proxy that forwards the connections to the current backend
            proxy_pass http://failover_backend;

            # default proxy headers
            proxy_set_header Host $host;
            proxy_set_header X-Real-IP $remote_addr;
            proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        }
    }

Build

Simply run:

go build