A reusable Django app that allows you to display a dashboard with any number of widgets to show any data you care about. The widgets are updated via socket.io, so you never need to refresh your dashboard.

You need to setup a socket.io-server that accepts incoming subscriptions and sends out broadcasts when widgets need an update. For this we have created https://github.com/bitmazk/django-socketio-messenger

TODO: Write a blog post about how to set this up on Webfaction.

You need at least the following packages in your virtualenv:

• Django 1.4.3
• South
• django-libs
• django-load
• requests
• django-socketio
• lockfile

To get the latest stable release from PyPi:

$ pip install django-metrics-dashboard

To get the latest commit from GitHub:

$ pip install -e git://github.com/bitmazk/django-metrics-dashboard.git#egg=metrics_dashboard

Add the app to your INSTALLED_APPS:

INSTALLED_APPS = [
    ...
    'metrics_dashboard',
]

Hook the app into your main urls.py:

urlpatterns += patterns('',
    ...
    url(r'^dashboard/', include('metrics_dashboard.urls')),
)

Run the south migrations to create the app's database tables:

$ ./manage.py migrate metrics_dashboard

Default: True

When you set this to false, anyone can access the dashboard. If you are displaying sensitive metrics, you might want to leave this at True.

Default: No default, you have to set this.

Set this to the API endpoint of your django-socketio-messenger installation. A valid value should look like this:

http://<HOST>[:<PORT>]/broadcast_channel/

Depending on your setup you might or might not need to specify a port.

We need this because all messages going through socketio must be sent from the same process. However, this app needs to broadcast messages from an admin command which get's executed from a cron job, therefore that command would be a different process than your wsgi process. As a simple (silly and hackish) solution we created django-socketio-messenger which is really just another mini Django app that functions as your socket.io server. Therefore this app would send HTTP requests to your django-socketio-messenger which then would broadcast those messages to your connected socket.io subscribers.

For now: Install it and go visit the URL :) More features coming soon.

• See https://github.com/bitmazk/md-pypispy-users as an example.
• Create a new Django app. Per convention, you should call your app something like md_yourwidgetname. This way we can easily search PyPi for md_ and will find all widgets that have been published.
• Give it a file dashboard_widget.py
• Implement your widget. It should inherit DashboardWidgetBase
• Your widget needs the following implementations:
  • a template_name attribute, just like any Django view
  • sizex & sizey attributes that define the widget size
  • an update_interval attribute in seconds.
  • a get_context_data method. It should return a dictionary of template context variables
  • a update_widget_data method. It should get data from a 3rd party API and save it to the widget's model. Then it should send a message to the widget's socket.io channel so that the subscribed browsers know that the widget has new data and needs an update.
• Register your widget with the dashboard_widget_pool.

Example dashboard_widgets.py:

from metrics_dashboard.widget_base import DashboardWidgetBase
from metrics_dashboard.widget_pool import dashboard_widget_pool

class DummyWidget(DashboardWidgetBase):
    """This widget is used by the tests."""
    template_name = 'dashboardwidget_dummy/dummy_widget.html'
    sizex = 2
    sizey = 1
    update_interval = 60

    def get_context_data(self):
        return {
            'value': 'Foobar',
        }

    def update_widget_data(self):
        # TODO: add example implementation here.

dashboard_widget_pool.register_widget(DummyWidget)

If you want to contribute to this project, please perform the following steps:

# Fork this repository
# Clone your fork
$ mkvirtualenv -p python2.7 django-metrics-dashboard
$ pip install -r requirements.txt

$ git co -b feature_branch master
# Implement your feature and tests
$ git add . && git commit
$ git push -u origin feature_branch
# Send us a pull request for your feature branch

If you want to contribute to this project you can run the tests without setting up a Django project. Just clone this repository and execute the runtests.py:

$ ./metrics_dashboard/tests/runtests.py

Sometimes a new feature needs new South migrations, in this case you should do the following:

$ rm db.sqlite
$ ./manage.py syncdb --migrate
$ ./manage.py schemamigration metrics_dashboard --auto

If you have questions or issues, please open an issue on GitHub.

If we don't react quickly, please don't hesitate to ping me on Twitter (@mbrochh)