rynge/einsteintoolkit repository overview

⁠Building and using Einstein Toolkit Docker containers

⁠Note for Open Science Grid

This is a repo forked from https://github.com/eschnett/einsteintoolkit-docker⁠ , and modified for use as a Singularity image under the Open Science Grid OSGVO.

⁠Overview

Docker makes it easy to set up a system configuration once, and then use it many times on different systems. Docker has "images" and "containers". An "image" is a snapshot of a system that is ready to be used. Running an image, i.e. running an application that is installed on an image, creates a "container" holding that running application. When the container stops, it can either be discarded, or can be turned into a new image. The original image is not modified by running it.

Thus images are static. For example, a freshly installed Ubuntu system is available as image. Below, we create a ready-to-be-used Einstein Toolkit install as image.

Containers, on the other hand, are dynamic, they correspond to running Unix processes. Each Einstein Toolkit simulation runs in a container.

Containers can interact with the host system. To

Finally, images can be easily be built from recipes called "Dockerfiles". A Dockerfile is essentially a shell script that contains all necessary commands to create the image from scratch.

⁠How to set up and run

docker build -t 'einsteintoolkit' .

IMAGE=f6477d58b757 docker commit $IMAGE et-sourcetree

docker run $IMAGE build --debug docker run -v Cactus:/home/scientist/Cactus