UserActionsLogging

An example app for user actions' logging. To run it, you need docker compose.

How to run it?

Create the env files. There are three:

analytics_app/.env

SECRET_KEY='anythingwithnumbers1231'
# If you change the ports in the docker compose you should change them here too
CELERY_BROKER_URL='redis://redis:6379/0' 
CELERY_RESULT_BACKEND='redis://redis:6379/0'
# Keep as secret. If you do not want to use the postgres container and want to use a local db,
# you have to check what the host direction should be.
DB_USER=user
DB_PASS=user_pass
DB_NAME=example
DB_HOST=db
DB_PORT=5432
# There's no need to change these but if you do make sure you 
# apply that change to the other files too
HOST='0.0.0.0'
PORT=5000

study_app/.env

# There's no need to change these but if you do make sure you 
# apply that change to the other files too
HOST='0.0.0.0'
PORT=5000

postgres.env

# These should be the same as in analytics_app
POSTGRES_USER=user
POSTGRES_PASSWORD=user_pass
POSTGRES_DB=example

Create and run the containers
```
docker-compose build
docker-compose up
```
Create the db

docker-compose run analytics_app python manage.py create_db

Architecture

There are 5 services running

1. Study app

The app that you want to test. If you want to use this as template, this folder should have the app that you are studying. The only file you should keep is the UserLogger.js to log the actions in your app.

2. Analytics app

The API that will receive the actions that you are logging. This app will create the celery tasks and put them in redis.

3. Celery

The task broker that will execute the data processing. This app is the same analytics app but running with another interpreter. This process will transform the raw actions from the user interface and put them in a database that you can use to analyze the results. This code should change according to the analysis you want to make, you can add other models (=other tables in the database) to save the data in the easiest format.

4. Redis

The queue that will have all the tasks that will be processed.

5. DB

A postgres database that will store the processed data. In this example it only has two tables: rawActions and PositionsScren. The rawActions table has the actions as they arrived from the user interface. The table PositionScren is storing the mouse position for each action with a timestamp. This data could be used to understand where on the screen the user is interacting and also to know how the user moves around the app.

indonoso/UserActionsLogging