There are two ways of running the VFrame Search application: you can use Docker, or you can install everything by hand. Following setup, please refer to steps 2-4 in the (README.md)[../README.md].
Dependencies can also be installed on your system manually.
conda env create -f environment.yml
conda activate vframe_search
We developed the frontend using Node v10 (LTS). We recommend using nvm to manage your local Node installation.
npm install
npm run build
Log into MySQL as root: mysql -u root and create a database. Corresponding settings should be put in the file .env in the root directory of the repo. Please see the file .env-sample for a template.
CREATE USER 'vframe_search'@'localhost' IDENTIFIED BY 'a very secure password';
CREATE DATABASE vframe_search;
GRANT ALL PRIVILEGES ON vframe_search TO 'vframe_search'@'localhost';
The Redis server can be run as a daemon, or manually using screen(1). Run the Celery command from inside cli/.
python `which celery` worker -A app.tasks.celery --loglevel=info -E
redis-server /usr/local/etc/redis.conf
With socket support:
python cli_flask.py socket
Without socket support (background indexing disabled):
python cli_flask.py run
Then open http://127.0.0.1:5000/