# The Littlest Jupyter Hub
Run a TLJH installation in a docker container.

## Build the image
Run:

```sh
$ docker build -t tljh-dev .
```

Or, if you have `make` installed:

```sh
$ make img
```

It'll download all its required dependencies during the build, it doesn't require anything from the context directory besides the `Dockerfile`.


## Start the container
Run the container:

```sh
$ mkdir -p data
$ docker run -d --rm --name=tljh-dev -p 12000:80 --privileged tljh-dev
```

Or, again, if you have `make` installed:

```sh
$ make start
```


### TO-DO
These details still need to be worked out:
 * Create a persistent mountpoint for all notebooks and custom configs
 * Speed up startup, by creating the user-environment during build time instead of doing it during runtime
 * Same as above for the systemd configuration (setting up services, 'daemon-reload', etc)