repo2docker image

Enterprise Applicable to Enterprise Edition

Community Applicable to Community Edition

PrimeHub could support notebook images built from repo2docker tool, if an images had

• sudo installed
• jovyan user is 1000 uid
• notebook started in Safe Mode

Please install repo2docker tool, before you get started.

Introduce repo2docker

repo2docker is a tool to simplify building a jupyter-notebook based image without giving a Dockerfile. We could put configuration files to make it working with repo2docker.

Build a PrimeHub compatible image

We will use binder-examples to show how to make it compatible with PrimeHub

git clone https://github.com/binder-examples/tensorboard

In order to make it work on the PrimeHub, we need to add sudo package. It needs an apt.txt file with this content:

sudo

After adding sudo, it is time to build an image with jupyter-repo2docker

• set user with --user-name jovyan
• set user uid with --user-id 1000
• configure the docker image tag by --image

jupyter-repo2docker --no-run --user-name jovyan --user-id 1000 \
--image infuseai/tensorboard-repo2docker-example .

If the command will invoke docker to build an image, you could push the image to docker repository and register it on the PrimeHub.

Launch notebook

Users should start a notebook built from repo2docker in Safe Mode, because it might have installed libraries in the $HOME.

---

Example: RStudio binder

We could follow steps to build a Jupyter Notebook with RStudio: https://github.com/binder-examples/r

1. Ensure repo2docker installed
2. Clone the source code and add sudo to apt.txt
3. Build image with arguments --user-name jovyan --user-id 1000
4. Add image to PrimeHub and start in Safe Mode

Check repo2docker

Execute jupyter-repo2docker command to see if any output from the command

$ jupyter-repo2docker --help | head -10
usage: jupyter-repo2docker [-h] [--config CONFIG] [--json-logs]
                           [--image-name IMAGE_NAME] [--ref REF] [--debug]
                           [--no-build]
                           [--build-memory-limit BUILD_MEMORY_LIMIT]
                           [--no-run] [--publish PORTS] [--publish-all]
                           [--no-clean] [--push] [--volume VOLUMES]
                           [--user-id USER_ID] [--user-name USER_NAME]
                           [--env ENVIRONMENT] [--editable]
                           [--target-repo-dir TARGET_REPO_DIR]
                           [--appendix APPENDIX] [--subdir SUBDIR] [--version]

If command not found, please install it by pip install jupyter-repo2docker

command not found: jupyter-repo2docker

Update configuration

Clone the source code

git clone https://github.com/binder-examples/r.git

In the source, create apt.txt and add sudo in a line

$ git status
On branch master
Your branch is up to date with 'origin/master'.

Untracked files:
  (use "git add <file>..." to include in what will be committed)

    apt.txt

nothing added to commit but untracked files present (use "git add" to track)

Verify

$ cat apt.txt
sudo

Build image

Build a image with the tag infuseai/r in the source directory (.)

jupyter-repo2docker --no-run --no-clean --user-name jovyan --user-id 1000 --push --image infuseai/r .

Add Image to PrimeHub

After the image taggedinfuseai/r is pushed, add it to the PrimeHub and start Notebook with the image in Safe Mode. Initially, it enters a general Jupyter Notebook.

For the RStudio environment, we must add the following at the end of the URL: ?urlpath=rstudio according to the README.

Regarding accessing RStudio, the current URL looks like

https://example.primehub.io/user/<username>/lab

Replace the suffix lab with rstudio to be like URL below and navigate.

https://example.primehub.io/user/<username>/rstudio

Here the RStudio is