Production deployment

Preparing production Docker images

Production deployment requires pre-build production Docker images to be available. Image names defined by the following env variables:

• DOCKER_IMAGE_BACKEND
• DOCKER_IMAGE_WORKER
• DOCKER_IMAGE_FRONTEND

Important

It is strongly recommended to use tagged Docker images in production deployment scenario to enforce stability! DOCKER_TAG env variable in .env file defines Docker image tag used during build and deployment.

In order to build these images and then push them to Docker image registry (e.g. Docker Hub), run the following command in your development environment:

make build-push

This operation should be done each time you are planning to release new production version.

Prerequisite

Make sure that the following tools installed globally on your machine:

• Git
• GNU Make
• Docker
• Docker Compose

Installation

As soon as Docker images are build and published to a registry, one can deploy production version of histoCAT by running the following commands on a production server:

Clone the histoCAT repo

git clone https://github.com/BodenmillerGroup/histocat-web.git

Later on you can update the repo by running:

git pull

Rename .env.template file in the root repo folder to .env and set all environment variables values accordingly to production configuration. Usually it should be done only once as .env file is ignored by Git and won't be overwritten by git pull command.

Warning

Please check that env variable DOMAIN is properly set in .env file on production server before deployment!

Deploy production Docker containers

make deploy-production

Info

Production deployment doesn't support debugging and hot-reloading because it is compiled to minimize bundle size and optimize performance.