AWE

AWE 2.0

AWE

A light weight workflow manager for scientific computing at scale that:


AWE is actively being developed at github.com/MG-RAST/AWE.

You can use AWE simultaenously on Clouds, Clusters and HPC systems with dozens, hundreds or thousands of nodes to run tens of thousands to hundreds of thousands of individuals workflows.

AWE Quickstart

This assumes that you have docker and docker-compose installed and curl is available locally.

Start AWE services

source ./init.sh
docker-compose up

This will start AWE server, one AWE worker, Shock object store and corresponding MongoDB containers. Don’t forget to later docker-compose down and do not forget, by default this demo configuration does not store data persistently.

Submit a job for simple workflow

This example consists of CWL workflow that takes a PDF file as input, extracts all words and generates a visual wordcloud. Execute the followin command in a new terminal from within the AWE folder:

 ./awe_submit.sh -w test/tests/pdf2wordcloud.cwl -j test/tests/rules-of-acquisition.job.cwl -d tmp

View result

open image with wordle/ preview here

Documentation

repository description link
AWE monitor UI for the AWE server github.com/MG-RAST/awe-monitor
Shock object store github.com/MG-RAST/Shock
Skyport2 demo environment using docker-compose github.com/MG-RAST/Skyport2

Papers to cite