AHM (Python)

AHM is a shortcut for Application for hypertension monitoring. Yes, I know, it is not the best name, but I like it.

The backend for this project was written in Python, there is another variant for AHM I'm planning to implement in Golang, the repository will be located in:

• https://github.com/jorlugaqui/ahm-serverless

And actually, I wrote about it in:

• https://medium.com/swlh/deploying-a-go-function-on-google-cloud-platform-9b79dd9d99aa

I have also recorded a workshop in which I have explained and shown the implementation's details for this Python variant. You could find it in:

• https://ed.team/cursos/flask-js (Spanish only - Let me know if you are interested in an English version)

Motivation

This application aims to serve as a measurement tracker for people who suffer high blood pressure issues, like me.

Architecture

The application follows a monolithic - single repository approach, with a three-tier architecture (data-logic-client).

There are four components in this application:

• api: The api component exposes a REST API which allows users to add and query blood pressure's measurements.

• aggregator: The aggregator lives within the API component and it is a standalone Python script thought to be run as a cron. It moves the measurements to the reporting entity.

• client: The client (web folder) is an independent vue JS application, which offers a form and a report page. Independent here does not mean it is a microservice.

• db: A document-based database (mongoDB).

DB design / Class model

There are three main entities that are related to each other as follow:

REST API

The API exposes the following endpoints:

HTTP METHOD	URI	Response
POST	/v1/measurement	201
PATCH	/v1/measurement/	200
GET	/v1/measurement/	200
GET	/v1/report/<time_period>	200

Examples:

• curl --location --request GET 'http://127.0.0.1:5000/v1/measurements/5ecbd76781201d2624f9583c'

• curl --location --request GET 'http://127.0.0.1:5000/v1/report/2020-05-31'

• curl --location --request POST 'http://127.0.0.1:5000/v1/measurements' \ --data-raw '{ "dia": 30, "pul": 66, "sys": 77 }'

• curl --location --request PATCH 'http://127.0.0.1:5000/v1/measurements/5ed329d0502dbd286623fd77' \ --data-raw '{ "sys": 1 }'

Business requirements

If you are curious, you can find the application's requirements in https://trello.com/b/SRSfqPB7/ahm.

Running the application in local

The application can be run locally using docker-compose. If you check the docker-compose file you will notice that the services are mapped to the main components in the architecture:

• mongo: The db component
• api: The api component, which covers the aggregator.
• web: The client component, which is a vue JS app.

Tech requirements

• Docker and docker-compose
• Git

Creating your secrets

In order to have the containers up and running, a set of env variables needs to be defined. You can do so by creating a .env file and adding the following entries:

MONGO_INITDB_ROOT_USERNAME=root
MONGO_INITDB_ROOT_PASSWORD=root
API_ENV=development
API_HOST=0.0.0.0
API_APP=main.py
DB_NAME=ahm
DB_HOST=mongo
VUE_APP_API_HOST=http://localhost:5000
WEB_ENV=development

Keep in mind those values should be only used for local development.

Running the database

docker-compose up mongo

You can also run a web UI interface (mongo-express) with:

docker-compose up mongo-express

Bear that the first time will take a while since it needs to download the images.

Defining a default user

For now, there is not support for authentication. Therefore, a default user is required. Once the mongo container gets working, you can create a user document like (in user collection):

{
	"email": "john@gmail.com",
	"name": "John",
	"surname": "Wick"
}

Keep in mind that the email you set is the email you need to add to the source code for filtering. This is very inconvenient, but authentication will come soon.

Running the API

docker-compose up api

You may want to build the image first with:

docker-compose build api

Running the UT locally for the API

docker-compose run api sh -c "pip install -r requirements_ci.txt && pytest --cov=. --cov-config=.coveragerc"

Running the client

The client requires a bit more patience for having it running in local with docker. First, let's build the image:

docker-compose build web

Second, let's install the requirements from a container

docker-compose run web npm install

Third, let's run the client:

docker-compose up web

Running the application in production

This application is not ready for production. But it eventually will.

Upcoming features

• Authentication
• More elaborated client
• Makefile for deployment (started)
• Unit testing for the API (started)
• Unit testing for the client
• Add CI (started)
• Add CD