Docker compose service recommender

[iamlax]

This adds containerization to the service recommender iis service

LESS INSTANCES MEANS CHEAPER COMPUTING COMPARED TO CURRENT SETUP.

This is a temp instance that i ran the docker-compose on to test the recommendation service

Application: http://34.73.211.38/

Recommender Endpoint: http://34.73.211.38:8128/api/recommendation/laxman_20@hotmail.com

Data Model Endpoint: http://34.73.211.38:8060/api/model

Temporary Account:

User: laxman_20@hotmail.com Pass: 123456789

READ BEFORE TESTING

1. To verify that the recommender is working make sure u have the key.json as well.
2. Copy the "# -- Production Recommendation Service ----" section from .env_local to .env on server Replace the fields to match ur instance for docker local it'll be something like

# -- Production Recommendation Service ----
RECOMMENDATION_SERVICE_PROTOCOL=http://
RECOMMENDATION_SERVICE_DOMAIN=192.168.99.100
RECOMMENDATION_SERVICE_PATH=api/recommendation
RECOMMENDATION_SERVICE_PORT=8128

Use scripts run_docker_dev.sh and run_docker_prod.sh to run application.

To test docker dev:

Run script run_docker_dev.sh. (make sure you are in a docker environment)

• Navigate to docker-machine ip on port 3000. (eg 192.168.99.100:3000).
• Navigate to docker-machine ip on port 8060/api/model. (eg 192.168.99.100:8060/api/model).
• Navigate to docker-machine ip on port 8128/api/recommendation/USER_EMAIL. (eg 192.168.99.100:8128/api/recommendation/user1@users.ca).

Test app as usual.

To test docker prod:

Run script run_docker_prod.sh. (make sure you are in a docker environment)

• Navigate to docker-machine ip on port 3000. (eg 192.168.99.100:3000).
• Navigate to docker-machine ip on port 8060/api/model. (eg 192.168.99.100:8060/api/model).
• Navigate to docker-machine ip on port 8128/api/recommendation/USER_EMAIL. (eg 192.168.99.100:8128/api/recommendation/user1@users.ca).Test app as usual.