dashboards.ly will be available in the next On-Premise release of plotly. In the meantime, here are some instructions to run your own version of dashboards.ly on your own servers.
1 - Clone the repo
$ git clone https://github.com/plotly/dashboards.ly.git
$ cd dashboards.ly
2 - Install dependencies into a python 2.7 virtual environment
$ pip install virtualenv
$ virtualenv venv
$ source venv/bin/activate
$ pip install -r requirements.txt
3 - Install postgres
4 - Set environment config variables
There are multiple configuration settings: "ProductionConfig", "TestingConfig", and "DevelopmentConfig".
These are set here: https://github.com/plotly/dashboards.ly/blob/master/dashboardsly/__init__.py
The variables in each of these environments are here: https://github.com/plotly/dashboards.ly/blob/master/dashboardsly/config.py
Here is an example of setting the environment variables
# Use the "production" environment
DASHBOARDSLY_PROD="true"; export DASHBOARDSLY_PROD
# Set the Postgres Database URI
DATABASE_URL="postgresql://chriddyp@localhost/chriddyp"; export DATABASE_URL;
# If you're running Plotly On-Premise,
# dashboards.ly can be configured to load graphs from your enterprise server
PLOTLY_API_DOMAIN="https://plotly.acme.com"
PLOTLY_DOMAIN="https://plotly.acme.com"
5 - Initialize the database
$ python init_db.py
6 - Run the server
$ gunicorn runserver:app