This page describes how to use Rasa Ephemeral Installer to install Rasa X.
Before You Start
You will need
sudo access on your machine to use REI for installation. The installer will prompt you for your password.
rasactl deploys Rasa X without a Rasa Open Source server. To connect a Rasa Open Source server,
see the Connect Rasa Open Source to Rasa X section.
You can learn more about
rasactl in the rasactl CLI docs section.
- Download and run Rasa Ephemeral Installer:
The command above will prepare the environment and install the following components:
rasactlto deploy Rasa :$ rasactl start∙∙∙ Ready!╭ Rasa X ───────────────────────────────────────────────╮│ ││ URL: http://agitated-goldberg.rasactl.localhost ││ Rasa X version: 1.1.0 ││ Rasa X password: rasaxlocal ││ │╰───────────────────────────────────────────────────────╯
rasactlwill deploy the latest stable release of Rasa X. To use the latest edge release of Rasa X, run the following instead:
rasactl start --rasa-x-edge-release
If you deploy Rasa X on a VM
rasactl will expose the Rasa X deployment on one of the node ports (
- Check the status of your deployment.
- Use the
rasactl opencommand to view Rasa X in a web browser. From there you can connect Rasa X to your bot's git repository.
Starting Rasa X with a Local Rasa Project
Starting a Rasa X deployment with a local Rasa project is supported for Rasa X / Enterprise ≥ 1.0.0.
The following instruction describes how to deploy Rasa X with a mounted local Rasa project.
kind. Using a local Rasa project is only supported with
kindas the k8s platform.
kindis installed along with other tools if you use the REI to install all components.
A local Rasa project. If you don't have one yet, there are two easy ways to create a Rasa project:
Starting Rasa X
- Start Rasa X in your local Rasa project directory:
You can use the
--project-path flag to define a Rasa project directory explicitly, in which case you don't have to run the command from your project directory.
- Check the status of your deployment - it should reflect your local project path: