Skip to content

Latest commit

 

History

History
42 lines (30 loc) · 1.34 KB

README.md

File metadata and controls

42 lines (30 loc) · 1.34 KB
Raw

content-recommendation

NLP-Based Content Recommendation Engine for the DP!

Development

Running the server requires Python 3 and Golang.

Loading the data

If you are using a new MongoDB instance, you can run go run main.go import to import articles from the DP API into the instance, and go run main.go update to add any recently published articles.

Setup

You will need Redis Stack running, which can be set up with the following commands:

docker pull redis/redis-stack-server
docker run -d --name redis-stack -p 6379:6379 redis/redis-stack-server:latest

You will also need to set the following environment variables

  • MONGO_URL
  • REDIS_URL
  • SITE_URL
  • CATEGORIES

If you are running the server for the first time, run pipenv install. Then you can run

pipenv shell
pipenv run python3 server.py --dev --reload

to start the server. The --reload flag can be omitted if you don't want to (re)load article data from MongoDB into Redis.

Deployment

First, set the following secrets with:

fly secrets set MONGO_URL=...
fly secrets set REDIS_URL=redis://localhost:6379

To deploy this service on Fly.io, run fly deploy --wait-timeout 1200 (the timeout is needed to load the articles into the Redis instance and index them before starting the server).