Aug 08 2013 0

Sending Push Notifications from RPi using Pushover

My Raspberry Pi has been humming along for quit some time now. Next thing to add to the project is a mechanism to send notifications when certain events occur (sudden increase/drop in temperature, etc.). Probably the simplest method for sending these type of notifications would probably be via email, but I don’t like it. I would rather implement some sort of push notifications.

I am already using Pushover to have IFTTT send notifications from several services to my mobile devices. So for me it makes sense to use the Pushover service. The Pushover service itself is free (no monthly fees), however the client applications are priced at $4.99 (iOS or Android).

You might also use Twitter for instance. This comprehensive article describes how to send tweets from your Raspberry Pi.

The FAQ contains sample code for numerous programming languages (PHP, Python, etc.) and even command line examples on how to send notifications using the Pushover service. To be able to use these code samples you will need to create an application.

You can Register your application on the following page over at Pushover and complete the form presented Register Pushover Application When your application is created a API Token/Key is presented.

NOTE: Every application allows you to send 7500 notifications per month (counter resets on the 1st of the month). If the case that 7500 notifications per month is not enough, you can buy 10.000 additional messages for $50 (or just create additional applications).

For Python there are two pieces of sample code provided on in the FAQ. Just make sure you pick the correct sample based on the version of Python you are running. For the Raspberry Pi I used the first piece of sample code, since mine is running version 2.7.3 of Python.


import httplib, urllib
conn = httplib.HTTPSConnection("")
conn.request("POST", "/1/messages.json",
    "token": "APP_TOKEN",
    "user": "USER_KEY",
    "message": "hello world",
  }), { "Content-type": "application/x-www-form-urlencoded" })

You can check your version of Python with the following command

python --version

With all of the code examples replace APP_TOKEN (line 5) with the token you received when creating your application and replace USER_KEY (line 6) with your own user key (presented your main page at Pushover).

Using the Python sample code I sent the following notification to my devices

iOS Android
Notification on iOS Notification on Android

The sample code only uses the following three required parameters;

  • token - your application’s API token
  • user - the user key (not e-mail address) of your user (or you), viewable when logged into our dashboard
  • message - your message

but there the API specifies also the following optional parameters;

  • device - your user’s device name to send the message directly to that device, rather than all of the user’s devices
  • title - your message’s title, otherwise your app’s name is used
  • url - a supplementary URL to show with your message
  • url_title - a title for your supplementary URL, otherwise just the URL is shown
  • priority - send as -1 to always send as a quiet notification, 1 to display as high-priority and bypass the user’s quiet hours, or 2 to also require confirmation from the user
  • timestamp - a Unix timestamp of your message’s date and time to display to the user, rather than the time your message is received by our API
  • sound - the name of one of the sounds supported by device clients to override the user’s default sound choice

So if you want to you can customize your notification even further, but for that is something for a later date since I first need to decide for which events I want to send out notifications and at what intervals the events need to be checked…

Previous post
Running it all on the Raspberry Pi (Part 3) - Performance Issue Resolved In my previous two post (Part 1 and 2) I described how I moved all of the processes for tracking the sensor data onto the Raspberry Pi. I ended the
Next post
Improving MySQL query performance Performance for the charts pages was getting worse and worse the more data was collected from all the sensors connected to the Raspberry Pi. It was
This blog is powered by Blot