Container API

The tagged versions of this data are also available on Kaggle as a Dinosaur Dataset.

This is a Docker image to do the following:

• Connect to a Google Storage Singularity Registry
• Compare the Containers in the Registry to the Local Record
• For new containers, pull and update the metadata

The metadata, which maintains the same path as Google Cloud storage, is served as a static json API on Github pages for researchers to take advantage of.

Build

docker build -t vanessa/container-api .

Generate or Update Database

export GOOGLE_STORAGE_BUCKET=singularityhub
# GOOGLE_APPLICATION_CREDENTIALS should be exported in environment

docker run --privileged -e GOOGLE_STORAGE_BUCKET \
                        -v $GOOGLE_APPLICATION_CREDENTIALS:/credentials.json \
                        vanessa/container-api

Interactive Development

docker run --privileged -it -e GOOGLE_STORAGE_BUCKET \
                            -v $PWD/data:/data \
                            -v $GOOGLE_APPLICATION_CREDENTIALS:/credentials.json \
                            --entrypoint bash \
                            vanessa/container-api

Run Locally

Note that you still have the requirements outlined in the Docker file! This is what @vsoch uses sometimes because it's all set up and easy to update the database :)

• spython=0.0.31
• singularity 2.4.5
• google-cloud-storage

export PATH=$PATH:../container-diff
export GOOGLE_STORAGE_BUCKET=singularityhub
# GOOGLE_APPLICATION_CREDENTIALS should be exported in environment
export API_DATABASE=$PWD/data

python generate.py