How to set up BinderHub

It’s possible to deploy your own BinderHub instance in order to serve on-the-fly Jupyter notebooks and computing environments. In order to do this, you’ll need four things:

1. Hardware that will run the BinderHub architecture and user instances 1. A Kubernetes cluster that is running and to which we have API access 1. A Docker registry that will host container images 1. A JupyterHub that’s configured to serve single-use nodes.

We’ll cover how to set up each of these below.

Choose your hardware

First you’ll need hardware on which we’ll run JupyterHub. This is most commonly a cloud computing platform such as Google Cloud, Microsoft Azure, or Amazon EC2. You can also deploy BinderHub on your own hardware if you so choose. <<<Add links to these deployments…maybe instructions to set up on their platforms as well…we should do this in z2jh and link here.>>>.

Set up a Kubernetes cluster

This is also platform-dependent, though it is relatively straightforward to set up kubernetes on most cloud computing providers. For instructions on how to do this, see the Zero to JupyterHub tutorial on setting up Kubernetes.

<<TODO: Configure JupyterHub to use binder>>