Does anyone have advice, know of resources or documentation for deploying Panel with Amazon Web Services (AWS)

New to Panel, but have so far loved the intuitiveness and wide range of functionality. I have become familiar with local deployment of a dashboard with Panel, but I am struggling to find any resources or documentation on how to deploy Panel via Amazon Web Services (AWS). Would greatly appreciate it if anyone could provide resources or a starting point.
Thank you.


Hi @dgaytanj

Welcome to the community. So far there are not a lot of walk throughs on deployment.

The starting point would be that you run your app(s) via either

‘panel serve’ on the command line and ‘.servable’ inside your file(s) or notebook(s)


‘Python’ on the command Line and ‘panel.serve’ inside your file’.

Start by verifying that the above works for you locally on your machine.

Please note that when you deploy you will run the same command. It’s not like flask or django where you have to use gunicorn as your server.

Then I would familiarize my self with the panel deployment docs.

There are some important settings to familiarize your self with like host, port, allowed hosts and number of processes. These can be specified on the command line, in file or as environment variables. Aws might have some requirements for their values.

If you are not familiar with deployment to aws, then I would google and find a deployment guide for Streamlit on aws . And follow that. Streamlit and panel both run on tornado server. So that is very similar and should be enough. You might also look for a guide on deploying bokeh to aws.

You might have to in aws enable for example web sockets and that your app should be always on.

For more advanced configuration you might have to consult the bokeh docs

Finally you might also study my docker files to see how I deploy in docker container to azure web apps for containers.

1 Like

And … @dgaytanj.

Please share your problems, solutions and the final steps. It would be so valuable for others.

You can do so in this post, as a separate blog post or as a PR to Panel.

Thanks :+1: