A headless face analytics platform built on top of dnfal library.

The following instructions are for installing Dnfas in Ubuntu 18.04 or newer. Installing on other system may work, but it has not been tested.

Before you are ready to run Dnfas, you will need to install the additional software on your computer. To install them, type:

sudo apt update
sudo apt install git python3-venv python3-dev pyhton3-apt

We recommend to run Dnfas within a newly created virtual environment for easier management. To create a new Python virtual environment and activate it, run the following commands within a directory of your choice:

python3 -m venv pyenv
source pyenv/bin/activate

Next, install python dependencies:

pip install wheel ansible

Next, change to a directory where you wish to install Dnfas and clone the repository:

git clone https://github.com/raikel/dnfas.git
cd dnfas

After that, you can install python dependencies by running:

pip install -r requirements.txt

Next step is to create a new database and user in Postgres. We provide a small script to quickly accomplish this task. Within Dnfas project root dir, type:

bash dbinit.sh \
    --pguser=${PG_USER} \
    --pgpass=${PG_PASS} \
    --dbname=${DB_NAME} \
    --dbuser=${DB_USER} \
    --dbpass=${DB_PASS} 

Where ${PG_USER} is a Postgres user with super-user privileges, ${PG_PASS} is the password of ${PG_USER}, and ${DB_NAME}, ${DB_USER} and ${DB_PASS} are the project database name, user and password, respectively. You can also set up the database by issuing the following commands in a Postgres session:

DROP DATABASE IF EXISTS ${DB_NAME};
CREATE DATABASE ${DB_NAME};
DROP USER IF EXISTS ${DB_USER};
CREATE USER ${DB_USER} WITH PASSWORD '${DB_PASS}';
ALTER ROLE ${DB_USER} SET client_encoding TO 'utf8';
ALTER ROLE ${DB_USER} SET default_transaction_isolation TO 'read committed';
ALTER ROLE ${DB_USER} SET timezone TO 'UTC';
GRANT ALL PRIVILEGES ON DATABASE ${DB_NAME} TO ${DB_USER};

To configure Dnfas runtime parameters, set the following environment variables:

A configuration file with all environments variables is also provided in the project. You can find it at deploy/dnfas.conf under the project root directory. To use, save it to a known location and edit its content, for example:

sudo cat deploy/dnfas.conf >> /etc/dnfas/dnfas.conf
nano /etc/dnfas/dnfas.conf

Then set the configuration variables to appropriated values.

With configuration parameters set up and within the project virtual environment, run the following command inside the project root directory to prepare the database:

python manage.py migrate

Next, to start the development server, type:

python manage.py runserver

Running Dnfas with the default server builtin with Django is a good way to start getting familiarized with the project. After that, however, you may want to run it with a higher performance server. Next, we describe how to setup Gunicorn and Nginx to serve Dnfas. Gunicorn will serve as an interface to Dnfas, translating client requests from HTTP to Python calls that our application can process. Nginx will be setup in front of Gunicorn to take advantage of its high performance connection handling mechanisms.