notice

This is documentation for Rasa X Documentation v0.37.x, which is no longer actively maintained.
For up-to-date documentation, see the latest version (1.1.x).

Version: 0.37.x

Installation Guide

Installing Rasa X on a server is easiest way to deploy your assistant to end users. Each server mode installation includes a Rasa Open Source container that serves your model, so when you deploy Rasa X, you deploy your assistant as well.

Rasa X can be installed in local mode or on a server using either Kubernetes/Openshift or Docker Compose. Consider which of the descriptions below best describe your situation and choose an installation method accordingly.

local mode or server mode?

Rasa X in local mode is helpful for sharing your assistant before you have a server set up. Once your assistant is up and running, you will want to deploy Rasa X to a server so that it’s available 24/7 and everyone on your team can use it to review conversations and annotate new training data.

Local Mode

If any of these describe your use case, this method is your best bet:

  • You are developing your Rasa assistant and want to chat with it in a UI and share it with others
  • You are exploring Rasa X for the first time, and just want to see how it works locally

You’ll need to use one of the other installation options if:

  • You are ready to deploy your assistant and make it available to end users

Server Quick-Install

This is the simplest and quickest way to install Rasa X on a server or cluster. If you’re not sure which approach to choose, try this one.

If any of these describe your use case, this method is your best bet:

  • You are installing Rasa X for use in production, and do not need much customization
  • You are exploring Rasa X for the first time, and want to see how it works on a server

You’ll need to use one of the other server installation options if:

  • Your system does not meet the requirements
  • You need customizations beyond those available for this method
  • You expect a lot of traffic to your bot (hundreds of concurrent users) and need to scale
  • You are installing Rasa Enterprise

Helm Chart

Installing Rasa X using the Helm chart is likely a good way to go if any of the following apply to you:

  • You are installing Rasa Enterprise
  • You expect a lot of user traffic (hundreds of concurrent users)
  • You need to customize things beyond what the quick-install script allows, for example:
    • You have a custom NLU component
    • You need to pull custom images for anything except the action server
    • You need to customize the database

You’ll need to use Docker Compose instead if:

  • You do not have and cannot acquire a cluster environment that meets the requirements

Docker Compose

Installing Rasa X on a server using Docker Compose would be a good choice if:

  • You can’t or don’t want to use Kubernetes/Openshift
  • You need more customization than the quick-install script provides
  • You are not expecting large amounts of user traffic (i.e. hundreds of concurrent users)

An install script is available for servers that meet the hardware and OS requirements. For other operating systems, you can install Rasa X via Docker Compose manually.