graph-notebook

This Dockerfile builds an aws/graph-notebook environment. It provides an easy way to interact with graph databases using Jupyter notebooks.

Apart from the required dependencies for running the aws/graph-notebook the Dockerfile also installs the aws/neptune-python-utils module together with a helper function that allows notebooks to connect to Neptune using Python and AWS IAM credentials.

Using neptune-python-utils

Run the docker image passing the environment variables corresponding to your "AWS variables", Neptune's host and Neptune's port:

docker run --rm -ti -p 8888:8888 \
    -e AWS_ACCESS_KEY_ID  -e AWS_SECRET_ACCESS_KEY -e AWS_SESSION_TOKEN \
    -e AWS_REGION="eu-west-1" \
    adevinta/graph-notebook-docker:v1.0.0

Alternatively you can share your local notebooks using the -v flag, for instance:

docker run --rm -ti -p 8888:8888 \
    -v $PWD:/notebooks \
    -e AWS_ACCESS_KEY_ID  -e AWS_SECRET_ACCESS_KEY -e AWS_SESSION_TOKEN \
    -e AWS_REGION="eu-west-1" \
    adevinta/graph-notebook-docker:v1.0.0

Then, execute the following code in a Notebook served from the container:

from neptune_helper import get_neptune_iam_connection
from gremlin_python.process.anonymous_traversal import traversal
conn = get_neptune_iam_connection("neptune.example.com", 8182)
g = traversal().withRemote(conn)

After that, you can use the g object to issue Gremlin queries:

g.V().limit(10).valueMap().toList()

Python dependencies

Both direct and transitive dependencies must be pinned. In order to do that we use pip-compile. After modifying the file requirements.in, you must run the following command to update requirements.txt:

script/pip-compile requirements.in requirements.txt

Versioning

We use semantic-versioning for releases. Each release in git has its corresponding tag in this dockerhub repository.

Contributing

This project is in an early stage, we are not accepting external contributions yet.

To contribute, please read the contribution guidelines in CONTRIBUTING.md.