Skip to content

Latest commit

 

History

History
130 lines (101 loc) · 4.3 KB

quick_start.md

File metadata and controls

130 lines (101 loc) · 4.3 KB
Raw

Quick start

System dependencies

Cloning and setting up git submodules

Clone this repo, and pull the submodules.

git clone https://github.com/cancerDHC/ccdh-terminology-service.git
cd ccdh-terminology-service 
git submodule update --init --recursive

Because the PDC json files are under Git LFS (Large File Storage), it must be installed in the repository. Follow these instructions for installing Git LFS on your machine. Then, install it in the repo via git lfs install. Then, pull the content with git lfs.

cd crdc-nodes/PDC-Public/documentation/prod/json
git lfs pull --include ./*.json
cd -  # to return to root diriectory of repo

Using Docker

Seeding Neo4j vlumes

This is the prefered approach to set up the service. You need to have docker and docker-compose installed on your system before the set up.

First copy the TCCM NCIT Turtle RDF to the docker import directory for neo4j.

cp data/tccm/ncit-termci.ttl docker/volumes/prod/neo4j/import
cp data/tccm/ncit-termci.ttl docker/volumes/test/neo4j/import

Environmental variables

In the root directory, Create a file called .env.prod file with required environment variables.

The .env.prod file should contain the following:

NEO4J_USERNAME=neo4j
NEO4J_PASSWORD=<password>
NEO4J_HOST=ccdh-neo4j
NEO4J_BOLT_PORT=7687
REDIS_URL=redis://ccdh-redis:6379
USER_ACCESS_TOKEN=<token>

Choose and <password>. As for the <token> to put under USER_ACCESS_TOKEN, this is used for GitHub workflow integration with the CCDH Model repository. If you have access to that repository, you should use a GitHub personal access token and set USER_ACCESS_TOKEN to that. The port, host, and 'bolt uri' have been auto-filled for you, but these are configurable if you want to change them.

By default, the importer will pull the CRDC-H YAML from the main branch of the ccdhmodel GitHub repo. If another branch is preferred, you can add this line in the .env file.

CCDHMODEL_BRANCH=ccdhmodel_branch_name_or_full_sha_commit_id

Finally, in addition to .env.prod, you'll also need a file called .env. This is due to some limitations in the docker setup. We want the docker-compose files to point to distinctly different .env files, but it also requires a generic file called .env, even if you tell it to look for another file. So you should run:

cp .env.prod .env

And finally, the current setup also expects an .env.test, even if you're just running a single instance of the docker containers. This really shouldn't be needed if runnin a single instance, but for now it is needed. So finally, run:

cdp .env.prod .env.test

Running containers

Then run the docker-compose build to build the images

make deploy-local

After the docker containers are up, log onto the ccdh-api container and load data.

docker exec -it ccdh-api /bin/bash
python -m ccdh.importers.importer

After data is loaded, the server will be running on a local port at 7070. You can visit http://localhost:7070.

Useful tools

You may find it useful to also install the following:

  • Docker desktop
  • Neo4j desktop / Neo4j browser

Debugging locally

If you want to debug locally, you should run the CCDH-API outside of docker. That way, if you set breakpoints in your IDE, it will stop at them.

Step 1: .env.dev 1.1. Create an .env.dev file. It can look like just like your .env.prod file, but with the following change:

NEO4J_HOST=localhost

1.2. Then, do: cp .env.dev .env

Step 2: Run From within your IDE's debugger, have it run the following command:

uvicorn ccdh.api.app:app $$ROOT_PATH --host 0.0.0.0 --port 8000