A minimal repository for running Apache™ Hadoop® on binderhub instances.
The tutorial for downloading, starting it in pseudo-distributed mode, running a map-reduce job, etc. is presented in the running-hadoop.ipynb. For a better visualization of the notebook please see the running-hadoop.ipynb via nbviewer.
- Launching this repository on a binderhub instance now
- Launching this repository locally using repo2docker
- Contributing
- License
Launching this repository on a binderhub instance now
Binderhub uses repo2docker for launching and serving the computational environments. The list below contains the known binderhub instances.
The list below provides the badges and the hyperlinks for launching Jupyter/Jupyter Lab in one of the binderhub instances.
Launching this repository locally using repo2docker
You can also launch this repository locally using Docker and repo2docker. Please refer to this link for installing Docker and this link for installing repo2docker.
The command below launches a container on port 8888. It also creates a Docker volume that maps the user's home into the container (the local-home folder).
Before running it, make sure your local user is in the docker group. Please refer to this Docker documentation for more details. It is strongly advised to not to run the container as root.
Please also be aware that the --ip 0.0.0.0 directive will start a sever which will accept connections from any ip. For security purposes the --NotebookApp.token='dstoken1234567' directive forces the need of a token for accessing any interface. Use the dstoken1234567 to login or feel free to set a stronger token.
repo2docker -p 8888:8888 \
-v $(echo ~):$(echo ~)/local-home \
https://github.com/thedatasociety/binderhub-hadoop \
jupyter lab --ip 0.0.0.0 --NotebookApp.token='dstoken1234567'Each interface will be available at a specific path, as follows:
-
Jupyter: http://0.0.0.0:8888/tree
-
JupyterLab: http://0.0.0.0:8888/lab
See the repo2docker documentation for more details regarding the use of multiple interfaces.