AlarmClock is a REST API for controlling starting up and shutting down computers via Wake on LAN and winrm.


The original inspiration for this project was the desire to wake and shutdown a PC using SmartHome voice control. By running this service on your home network and adding a webhook to IFTTT, a user can use voice control to wake up and shutdown a machine.



In order to successfully use AlarmClock, a config.yml file must be created and completed. The configuration file has the following format:

broadcast: "<broadcast_address>"
    mac: "12:34:56:78:90:AB"
    username: "<username>"
    password: "<password>"
    mac: "12:34:56:78:90:AB"
    username: "<username>"
    password: "<password>"

It is highly recommended that vault be used to securely store usernames and passwords. See the section below for instructions on integrating AlarmClock with Vault.

The broadcast address for your home network can be obtained by running ipconfig Windows or ifconfig Linux/Mac from a terminal. Typically, this will be something similar to: The mac address for the machines to control wake/sleep can be found using the same commands.

Currently only a single broadcast is supported for all hosts. Also, only Windows support is available for shutdown. Username and password are not required in config.yml if shutdown functionality is not needed.

Vault Support

It is highly recommended to use Vault to securely store usernames/passwords for machines on your network. In order to use vault set the following environment variables before running AlarmClock:


Ensure your vault server is unsealed and the token provided is correct. AlarmClock attempts to lookup usernames/passwords under the secrets path secret/alarmclock.

It expects the usernames/passwords to have the following key/value format <hostname>_username and <hostname>_password. For example, for the config.yml example listed above the following key/values are expected.


Running AlarmClock

Via Docker

AlarmClock can be run via docker using the following command:

docker run -d -v <path_to_config.yml>:/config.yml --network="host" speyside/alarmclock:latest

It is required to run AlarmClock on the host network due to the requirement to send magic packets to mac addresses on the network to wake them.

Via Binaries

The AlarmClock binaries are available for most distributions and are available at the following location:

To run AlarmClock, pass the location of the config file to the binary via:

./alarmclock -f <path_to_config>/config.yml

Configuring Shutdown


Windows needs special configuration to allow remote management and shutdown. To enable shutdown follow the instructions below.

On the remote host, a PowerShell prompt, using the Run as Administrator option and paste in the following lines:

winrm quickconfig
winrm set winrm/config/service/Auth '@{Basic="true"}'
winrm set winrm/config/service '@{AllowUnencrypted="true"}'
winrm set winrm/config/winrs '@{MaxMemoryPerShellMB="1024"}'

Windows Firewall must be running to enable remote management. In addition the NIC used for shutdown, must be set as a private network (network discovery enabled).


TunnelFox is a python command line tool for port-Forwarding via SSH. Have you ever needed to access a port on a remote server, but the port is not accessible. If you have SSH access you can forward the port via ssh so that it is accessible locally.

For example, if I want to access port 8080 on a remove server, I can use TunnelFox to access that port locally. By running: tunnelfox new -s -p 8080. I can locally access hat is running on by running curl http://localhost:8080.


TunnelFox can be used to start, stop and manage existing tunnels.


To list existing tunnels run:

$ tunnelfox ls

1: 5050:5050
2: 8000:8000
3: 9000:9000

The format for the output is:

<remote_host> <locally_accessible_port>:<remote_port>

If a connection died (i.e. the connection was interrupted) it will be displayed in the output:

$ tunnelfox ls

1: 5050:5050
2: 8000:8000 (dead)
3: 9000:9000


To establish new tunnels use the new command.

Establish a new tunnel to forwarding remote port 8080 to local port 8080:

tunnelfox new -s -p 8080 -l 8080

A shorter version of forwarding remote port 8080 to local port 8080:

tunnelfox new -s -p 8080

Establish a new tunnel to forwarding remote port 443 to local port 8443:

tunnelfox new -s -p 443 -l 8443


To stop existing tunnels use the ls command to find their numbers then use the stop command:

tunnelfox stop 2


With pip installed, run

pip install tunnelfox

You can now run tunnelfox from the command line.


To contribute simply fork from GitHub and submit a Pull Request.


MIT License