When jupyter-panel-proxy is installed and you launch a Jupyter server (Notebook, JupyterLab or JupyterHub), a Panel server will be launched when you visit the /panel endpoint of the server. This will show an index of all applications being served, to launch a particular application visit the corresponding endpoint /panel/<name_of_file>.
The jupyter-panel-proxy is available from pip:
pip install jupyter-panel-proxy
and conda:
conda install -c pyviz jupyter-panel-proxy
The jupyter-panel-proxy provides the ability to configure the proxy server by declaring a jupyter-panel-proxy.yml in the directory the Jupyter server is being launched from. The yaml file may declare the following keys:
apps(list): A list of applications or glob patterns to servelauncher_entry(dict): A jupyter-server-proxy launcher entry specificationfile_types(list(str)): A list of file types to serve if no explicit apps list is providedexclude_patterns(list(str)): A list of glob/(fnmatch) patterns to exclude specific applicationsindex(str): The path to a Bokeh index templateautoreload(bool): Whether to automatically reload user sessions when the application or any of its imports change.static_dirs(list): A list of dicts mapping from server route to the static directory to be servedwarm(bool): Whether to execute scripts on startup to warm up the server.num_procs(int): Number of worker processes for an app. Using 0 will autodetect number of cores (defaults to 1)oauth_provider(str): The OAuth2 provider to use.oauth-key(str): The OAuth2 key to useoauth-secret(str): The OAuth2 secret to useoauth-redirect-uri(str): The OAuth2 redirect URIoauth_extra_params(dict): Additional parameters to the OAuth provider.oauth_jwt_user(str): The key in the ID JWT token to consider the user.