Modular Architecture - Technology stack illustrated

Docker (intro)

• Base principles of Docker (image vs. container,... ) and Docker compose

• DEMO: Using Docker to start a mssql database and restore (current) openIMIS backup

  • Configuring Microsoft Docker image in Docker-container yml file (mounting volumes,...)

  • Trigger image download and launch container:

    • docker-compose up imis-db

  • List running containers:

    • docker container ls

  • Connect to docker: 

    • docker exec -it openimis-poc_imis-db_1 /bin/bash

  • ... and, within the docker, restore database:

"All proxy" modularized frontend configuration

• From openimis react app, the plugin.js file to activate plugins

• Contributing to MainMenu with a simple entry point & fake module 

• In locations_proxy/index.js, main menu contribution et ProxyPage

• Building the front docker image:  Dockerfile, particularizing the node:8, adding the code and launching the webpack packaging for the front

• Launch of the front docker image from docker-compose and access via the gateway (transmit session to openIMIS)

Implementing the Location backend module

• Reverse of imis (mssql database) into django model:

python manage inspectdb > reverse/models.py

• Moving the tblLocation and tblPayer to own module and fine tuning the model (Foreign keys,...)

• python shell and navigation in model

• exposing Location via the django /admin console

• exposing the Location in json:api

• exposing the Location in dynamic rest api

Replacing the locations_proxy by react locations implementation

• Creating the Location react component and binding to rest api

• Displaying Payers info within the Location detail, without hard dependence between module:

  • The PayersOfLocationList.js contribution

  • The LOCATION_SELECTED redux event

Database switch

• Switching from managed = False to managed = True in django model

• Changing the table name convention (prefixing by module name,...)

• Clearing and regenerating migrations 

• Switching database and recreate the database structure

• Moving data from mssql to postgres

  • via CSV

  • via JSON... and JSONB column (via pgfutter)... some JSONB features

Module refactorings

• adding a uuid to models

• adapting the generated migration script to generate the uuid on existing database

Adopting FHIR as core data model

• Using fhirbase to generate the FHIR database schema

• Importing Locations into the FHIR tables... and adpatations in

  • the backend (django model)

  • the frontend