First time setup guide

Create the global docker network

docker network create my-app-network

Setup LSAAI Mock

Generate a self-signed certificate for the LSAAI nginx proxy using this command:

openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout ./lsaai-mock/configuration/nginx/certs/nginx.key -out ./lsaai-mock/configuration/nginx/certs/nginx.crt -addext subjectAltName=DNS:aai.localhost

Setup REMS

1. Generate jwks

cd rems
python generate_jwks.py
cd ..

2. Initialize the database

./rems/migrate.sh

3. Run the rems-app service

docker compose up -d rems-app

4. Log in to the REMS application (http://localhost:3000), so your user gets created in the database
5. Initialize the application

./rems/setup.sh

At steps 2, 3 and 5 it is important not to cd into the rems directory.

Run the rest of the services

docker compose up -d

Uploading a dataset to SDA

In upload.sh there is a simple script that uploads a dataset to SDA.

Prerequisites

• sda-admin and sda-cli in your $PATH
• access token from http://localhost:8085/
• C4GH public key generated (and printed on startup) in the storage-and-interfaces/credentials container
• s3cmd.conf file in the same directory as the script. You can just replace <access_token> in this example:

[default]
access_token = <access_token>
human_readable_sizes = True
host_bucket = http://localhost:8002
encrypt = False
guess_mime_type = True
multipart_chunk_size_mb = 50
use_https = False
access_key = jd123_lifescience-ri.eu
secret_key = jd123_lifescience-ri.eu
host_base = http://localhost:8002
check_ssl_certificate = False
encoding = UTF-8
check_ssl_hostname = False
socket_timeout = 30

Usage

1. Create a file named crypt4gh_key.pub and copy the key from the storage-and-interfaces/credentials container output log
2. Move all your data to a folder next to the script
3. Run ./upload.sh <folder_name> <dataset_name>
4. After successful upload, you should be able to fetch the data using:

token=<access_token>
curl -H "Authorization: Bearer $token" http://localhost:8443/s3/<dataset>/jd123_lifescience-ri.eu/<folder_name>/<file_name>

Running a workflow from the compute-web

1. Log in to http://localhost:4180
2. Select a workflow on the home page
3. Define the input directory as <dataset>/<user_id> (e.g. DATASET001/jd123_lifescience-ri.eu)
4. Define the output directory. This is the directory in the s3 inbox bucket (e.g., myWorklowOutput)
5. Click Run