rasahq/rasactl

rasactl deploys Rasa X / Enterprise on your local or remote Kubernetes cluster and manages Rasa X / Enterprise deployments.

Project Status

rasactl is not maintained anymore.

See Rasa X Community Edition - Deprecation Notice for more information.

rasactl

rasactl deploys Rasa X / Enterprise on your local or remote Kubernetes cluster and manages Rasa X / Enterprise deployments.

Features

• deploy Rasa X / Enterprise

  You can use rasactl to deploy Rasa X / Enterprise on your local machine or a VM in one of the major cloud providers.

  (check the Prerequisites section)

• upgrade Rasa X / Enterprise

  Upgrade/change configuration for an existing Rasa X deployment.

• stop/delete/start a running Rasa X / Enterprise deployment

  Manage the lifecycle of your deployment: you can stop, delete or start one of the Rasa X deployments managed by rasactl.

• connect a local Rasa Server to Rasa X / Enterprise

  You can use your local Rasa Open Source server along with Rasa X / Enterprise. rasactl will prepare configuration for Rasa OSS and Rasa X and run the Rasa Open Source server on your local machine.

  (requires kind and Rasa OSS installed locally)

• use a local Rasa project along Rasa X / Enterprise deployment

  Use your local Rasa project along with Rasa X / Enterprise deployment. The rasactl provides an easy way to use your local Rasa project along with Rasa X / Enterprise.

  This setup was previously referred to as "local mode" in older Rasa X versions.

  (requires kind and Rasa X >= 1.0.0)

Table of Contents

• rasactl
  • Features
  • Table of Contents
  • Prerequisites
  • Installation
    • Linux / macOS
  • Compatibility matrix
  • Before you start
  • Values File
  • Configuration
    • Environment variables
    • Configuration file
  • Global flags
  • Commands
    • The add command
    • The start command
    • The stop command
    • The delete command
    • The list command
    • The status command
    • The config use-deployment command
    • The connect rasa command
    • The auth login command
    • The auth logout command
    • The logs command
  • Enterprise Management Commands
    • The enterprise activate command
    • The enterprise deactivate command
  • Model Management Commands
    • The model delete command
    • The model download command
    • The model list command
    • The model tag command
    • The model upload command
    • Upload a model to Rasa X
  • Examples of usage
    • Run Rasa X / Enterprise with a local Rasa Server
    • Run Rasa X / Enterprise with mounted a local Rasa project
    • Upgrade Rasa X / Enterprise version
    • Deploy Rasa X in one of the public cloud providers
  • Development
    • How to run it?
    • Run unit tests
    • Kind cluster for developing purposes
  • Create a new release
  • License

Prerequisites

• Kubernetes cluster >= 1.14+

  or

• kind (for local mode)

(You can use the REI to install all required components on your local machine or a VM.)

Installation

Linux / macOS

• Binary downloads of rasactl can be found on the Releases page. You can manually install rasactl by coping the binary into your bin:

$ curl -L https://github.com/RasaHQ/rasactl/releases/download/0.0.24/rasactl_0.0.24_darwin_amd64.tar.gz -O
$ tar -zxvf rasactl_0.0.24_darwin_amd64.tar.gz
$ cp rasactl_0.0.24_darwin_amd64/rasactl /usr/local/bin/

• You can also install via brew:

$ brew tap rasahq/rasactl
$ brew install rasactl

Compatibility matrix

rasactl version	Helm chart version
>= 1.0.x	4.x
<= 0.5.x	3.x

Before you start

Below you can find several things that are good to know and keep in mind when you use rasactl.

• It is possible to configure multiple deployments with rasactl. A rasactl command will always execute an operation on a single deployment. Here is the order in which rasactl determines which deployment to use:

  1. A deployment name passed as an argument in CLI, e.g. rasactl status deployment-name, you can use rasactl help command to see usage example for a given command.
  2. rasactl checks if a .rasactl file exists in a current working directory. If so, the deployment defined in the file is used. This .rasactl file is created automatically when the rasactl start --project command is executed.
  3. rasactl checks if a default deployment is configured in the rasactl.yaml configuration file, if yes, then the default deployment is used. The default deployment can be set by using the rasactl config use-deployment command.
  4. If there is only one deployment, then it's used.

  You can use the rasactl list command to check which deployment is used as the current one.

  The rasactl delete command requires explicitly passing a deployment name as an argument.

• rasactl uses the rasa-x-helm chart to deploy Rasa X / Enterprise.

• rasactl deploys Rasa X / Enterprise without a Rasa Open Source server. It's up to you to connect Rasa OSS with Rasa X / Enterprise deployment.

• rasactl uses a Kubernetes context from the kubeconfig file, if you want to switch Kubernetes cluster you have to use kubectl or other tools that change the active context for the kubeconfig.

Values File

The rasactl uses the rasa-x-helm chart to deploy Rasa X / Enterprise, which means you can use the helm chart values to configure deployment. The rasactl enables template usage for the values file so that it's possible to use the Go template and Sprig function within the value file, e.g.

# values.yaml
rasax:
  podLabels:
    rasactl: "true"
    test_version: {{ env "RASACTL_TEST_VERSION" }}
    test_template: {{ coalesce 0 1 2 }}

Configuration

Environment variables

Name	Description
RASACTL_AUTH_USER	The username that is used to authorize to Rasa X / Enterprise
RASACTL_AUTH_PASSWORD	The password that is used to authorize to Rasa X / Enterprise
RASACTL_RASA_X_URL	Set Rasa X / Enterprise URL. By default, the URL is detected automatically, but if you use a custom configuration and you wanna define Rasa X URL explicitly you can use the env variable. The RASACTL_RASA_X_URL overrides Rasa X URL for all deployment.
RASACTL_RASA_X_URL_<DEPLOYMENT_NAME>	Set Rasa X / Enterprise URL for a given deployment, e.g. if a deployment name is my-deployment, then you can use the RASACTL_RASA_X_URL_MY_DEPLOYMENT environment variable to define the Rasa X URL for the my-deployment.
RASACTL_KUBECONFIG	Absolute path to the kubeconfig file (default "$HOME/.kube/config")
RASACTL_SKIP_DOCKER_VERSION_CHECK	Don't check if the Docker engine version is incompatible with rasactl. Default is false.

Configuration file

Below you can find an example of the configuration file and parameters that can be defined, by default configuration file is located in $HOME/.rasactl.yaml.

# Deployment name that is used as a current deployment (default).
# You can use the `rasactl config use-deployment` command to set the current deployment.
current-deployment: my-deployment

# Name of the kubeconfig context to use
kube-context: ""

# Absolute path to the kubeconfig file
kubeconfig: /home/user/.kube/config

Global flags

Below you can find global flags that can be used with every command.

Global Flags:
      --config string         config file (default is $HOME/.rasactl.yaml)
      --debug                 enable debug output
  -h, --help                  help for rasactl
      --kube-context string   name of the kubeconfig context to use
      --kubeconfig string     absolute path to the kubeconfig file (default "$HOME/.kube/config")
      --verbose               enable verbose output

Commands

Available Commands:
  add         add existing Rasa X deployment to rasactl
  auth        manage credentials for Rasa X / Enterprise
  completion  generate the autocompletion script for the specified shell
  config      modify the configuration file
  connect     connect a component (e.g. a Rasa OSS server) to Rasa X
  delete      delete Rasa X deployment
  enterprise  manage Rasa Enterprise
  help        Help about any command
  list        list deployments
  logs        print the logs for a container in a pod
  model       manage models for Rasa X / Enterprise
  open        open Rasa X in a web browser
  start       start a Rasa X deployment
  status      show deployment status
  stop        stop Rasa X deployment
  upgrade     upgrade Rasa X deployment

The add command

Adds existing Rasa X deployment to rasactl.

If you already have a Rasa X deployment that uses the rasa-x-helm chart you can add the deployment and manage it by rasactl.

Usage:
  rasactl add NAMESPACE [flags]

Examples:
  # Add a Rasa X deployment that is deployed in the 'my-test' namespace.
  $ rasactl add my-test

  # Add a Rasa X deployment that is deployed in the 'my-test' namespace and
  # a helm release name for the deployment is 'rasa-x-example'.
  $ rasactl add my-test --rasa-x-release-name rasa-x-example

Flags:
  -h, --help                         help for add
      --rasa-x-release-name string   a helm release name to manage (default "rasa-x")

The start command

The start command creates a Rasa X deployment or starts a stopped deployment if a given deployment already exists.

Usage:
  rasactl start [DEPLOYMENT-NAME] [flags]

Examples:
  # Create a new Rasa X deployment with an autogenerated name.
  $ rasactl start

  # Create a Rasa X deployment with a defined name.
  $ rasactl start my-deployment

  # Create a new deployment if there is already one or more deployments.
  # rasactl start --create

  # Create a Rasa X deployment with custom configuration, e.g the following configuration changes a Rasa X version.
  # All available values: https://github.com/RasaHQ/rasa-x-helm/blob/main/charts/rasa-x/values.yaml
  $ rasactl start --values-file custom-configuration.yaml

  # Create a Rasa X deployment with a defined password.
  $ rasactl start --rasa-x-password mypassword

  # Create a Rasa X deployment that uses a local Rasa project.
  # The command is executed in a Rasa project directory.
  $ rasactl start --project

Flags:
      --create                        create a new deployment. If --project or --project-path is set, or there is no existing deployment, the flag is not required to create a new deployment
  -h, --help                          help for start
  -p, --project                       use the current working directory as a project directory, the flag is ignored if --project-path is used
      --project-path string           absolute path to the project directory mounted in kind
      --rasa-x-chart-version string   a helm chart version to use
      --rasa-x-edge-release           use the latest edge release of Rasa X
      --rasa-x-password string        Rasa X password (default "rasaxlocal")
      --rasa-x-password-stdin         read the Rasa X password from stdin
      --rasa-x-release-name string    a helm release name to manage (default "rasa-x")
      --values-file string            absolute path to the values file
      --wait-timeout duration         time to wait for Rasa X to be ready (default 10m0s)

The stop command

The stop command stops a running Rasa X / Enterprise deployment. The Rasa X deployment and all its components will be scaled down to 0.

Usage:
  rasactl stop [DEPLOYMENT-NAME] [flags]

Examples:
  # Stop a Rasa X deployment with the 'my-deployment' name.
  $ rasactl stop my-deployment

  # Stop a currently active Rasa X deployment.
  # The command stops the currently active deployment.
  # You can use the 'rasactl list' command to check which deployment is currently used.
  $ rasactl stop

Flags:
  -h, --help   help for stop

The delete command

The delete command deletes a Rasa X / Enterprise deployment.

You can use the --prune flag to remove a namespace where Rasa X deployment is located.

Notice If you want to free resources, or temporarily you don't need to run Rasa X deployment, you can stop a Rasa X / Enterprise deployment instead of deleting it. Stopping the Rasa X deployment will free resources, but keep the current configuration.

Usage:
  rasactl delete DEPLOYMENT-NAME [flags]

Aliases:
  delete, del

Examples:
  # Delete the 'my-example' deployment.
  $ rasactl delete my-example

  # Prune the 'my-example' deployment, execute the command with the --prune flag deletes the whole namespace.
  $ rasactl delete my-example --prune

Flags:
      --force   if true, delete resources and ignore errors
  -h, --help    help for delete
      --prune   if true, delete a namespace with a project

The list command

List all deployments.

$ rasactl list
CURRENT	NAME         	STATUS 	RASA PRODUCTION	RASA WORKER	ENTERPRISE	VERSION
       	hopeful-haibt	Running	2.8.1          	2.8.1      	inactive  	0.42.0
*      	vibrant-yalow	Running	2.8.1          	2.8.1      	inactive  	0.42.0

The * in the CURRENT field indicates a deployment that is used as default. It means that every time when you execute rasactl command without defining the deployment name, the deployment marked with * is used.

A deployment is marked as CURRENT if:

• there is a .rasactl file that includes a deployment name in your current working directory. The file is automatically created if you run the rasactl start command with the --project or --project-path flag
• there is only one deployment
• you set the current deployment by using the rasactl config use-deployment command

The status command

Show the status of a deployment.

Usage:
  rasactl status [DEPLOYMENT-NAME] [flags]

Examples:
  # Show status for the 'example' deployment.
  $ rasactl status example

  # Show status for the 'example' deployment along with details.
  $ rasactl status example --details

Flags:
  -d, --details         show detailed information, such as running pods, helm chart status
  -h, --help            help for status
  -o, --output string   output format. One of: json|table (default "table")

Example output:

$ rasactl status vibrant-yalow
Name:                   	vibrant-yalow
Status:                 	Running
URL:                    	http://vibrant-yalow.rasactl.localhost
Version:                	0.42.0
Enterprise:             	inactive
Rasa production version:	2.8.1
Rasa worker version:    	2.8.1
Project path:           	/home/ubuntu/test

The config use-deployment command

Sets the current-deployment in the configuration file.

If you have multiple Rasa X deployments, and you are not in a project directory you have to explicitly define the deployment name during command execution. You can define a deployment that is used as a current one by using the rasa config use-deployment command.

Usage:
  rasactl config use-deployment DEPLOYMENT-NAME [flags]

Examples:
  # Set the 'example' deployment as the current deployment.
  $ rasactl config use-deployment example

Flags:
  -h, --help   help for use-deployment

The connect rasa command

Run a local Rasa Open Source server and connect it to a Rasa X deployment.

The command prepares a configuration that's required to connect Rasa X deployment and run a local Rasa server.

It's required to have the 'rasa' command accessible by rasactl.

The command works only if Rasa X deployment runs on a local Kubernetes cluster managed with 'kind'.

Usage:
  rasactl connect rasa [DEPLOYMENT-NAME] [flags]

Examples:
  # Connect Rasa Server to Rasa X deployment.
  $ rasactl connect rasa

  # Run a separate rasa server for the Rasa X worker environment.
  $ rasactl connect rasa --run-separate-worker

  # Pass extra arguments to rasa server.
  $ rasactl connect rasa --extra-args="--debug"

Flags:
      --extra-args strings    extra arguments for Rasa server
  -h, --help                  help for rasa
  -p, --port int              port to run the Rasa server at (default 5005)
      --run-separate-worker   runs a separate Rasa server for the worker environment

The auth login command

Log in to Rasa X / Enterprise.

auth login stores credentials in an external credentials store, such as the native keychain of the operating system.

The following external credential stores will be used:

• On macOS: Apple macOS Keychain Access
• On Linux: pass
• On Windows: Microsoft Windows Credential Manager

You can pass credentials via environment variables:

• RASACTL_AUTH_USER - username
• RASACTL_AUTH_PASSWORD - password

If the environment variables are used, credentials stored in a native keychain are not used.

Usage:
  rasactl auth login [DEPLOYMENT-NAME] [flags]

Examples:
  # Login to the 'my-deployment' Rasa X / Enterprise deployment.
  $ rasactl auth login my-deployment

  # Login to Rasa X / Enterprise (login to the currently active deployment).
  $ rasactl auth login

  # Provide a password using STDIN.
  # You can login non-interactively by using the --password-stdin flag to provide a password through STDIN.
  # Using STDIN prevents the password from ending up in the shell’s history.
  $ rasactl auth login --username me --password-stdin

Flags:
  -h, --help              help for login
  -p, --password string   password
      --password-stdin    read the password from stdin
  -u, --username string   username

Notice For Linux, pass is used as credential storage. pass must be installed and configured before you use the rasactl auth command. Below you can find an example of pass installation and configuration.

pass installation and configuration for Linux Ubuntu.

1. Install pass.

sudo apt-get install pass

2. Generate a GPG key.

$  gpg --gen-key
gpg (GnuPG) 2.2.19; Copyright (C) 2019 Free Software Foundation, Inc.
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.

gpg: directory '/home/ubuntu/.gnupg' created
gpg: keybox '/home/ubuntu/.gnupg/pubring.kbx' created
Note: Use "gpg --full-generate-key" for a full featured key generation dialog.

GnuPG needs to construct a user ID to identify your key.

Real name: rasactl
Email address:
You selected this USER-ID:
    "rasactl"

Change (N)ame, (E)mail, or (O)kay/(Q)uit? O
[...]
public and secret key created and signed.

3. Init pass.

$ pass init rasactl
mkdir: created directory '/home/ubuntu/.password-store/'
Password store initialized for rasactl

Now you can use rasactl auth on Linux.

$ rasactl ls
CURRENT	NAME             	STATUS 	RASA PRODUCTION	RASA WORKER	ENTERPRISE	VERSION
*       wonderful-gagarin	Running	2.8.1          	2.8.1      	inactive  	0.42.0
$ rasactl auth login
Username: me
Password:
Successfully logged.

Troubleshooting If you see Error: exit status 2: gpg: decryption failed: No secret key error you should export the following environment variable export GPG_TTY="$(tty)".

The auth logout command

Removes credentials from an external credentials store, such as the native keychain of the operating system.

Usage:
  rasactl auth logout [DEPLOYMENT-NAME] [flags]

Examples:
  # Remove access credentials (use the currently active deployment).
  $ rasactl auth logout

  # Remove access credentials for the 'my-deployment' deployment.
  $ rasactl auth logout my-deployment

Flags:
  -h, --help   help for logout

The logs command

Print the logs for a container in a pod. If the pod has only one container, the container name is optional.

Usage:
  rasactl logs [DEPLOYMENT-NAME] [POD] [flags]

Examples:
  # Choose a pod and show logs for it (use the currently active deployment).
  $ rasactl logs

  # Show logs from pod rasa-x (use the currently active deployment).
  $ rasactl logs rasa-x

  # Show logs from pod rasa-x for the 'my-deployment' deployment.
  $ rasactl logs my-deployment rasa-x

  # Display only the most recent 10 lines of output in pod rasa-x
  $ rasactl logs rasa-x --tail=10

  # Return snapshot of previous terminated nginx container logs from pod rasa
  $ rasactl logs -p -c nginx rasa

  # Begin streaming the logs from pod rasa-x
  $ rasactl logs -f rasa-x

Flags:
  -c, --container string   a container name
  -f, --follow             specify if the logs should be streamed
  -h, --help               help for logs
  -p, --previous           print the logs for the previous instance of the container in a pod if it exists
      --tail int           lines of recent log file to display. Defaults to -1 showing all log lines (default -1)

Enterprise Management Commands

You can manage an Enterprise license via rasactl.

manage Rasa Enterprise

Usage:
  rasactl enterprise [command]

Available Commands:
  activate    activate an Enterprise license
  deactivate  deactivate an Enterprise license

The enterprise activate command

Activate an Enterprise license.

Usage:
  rasactl enterprise activate [DEPLOYMENT-NAME] [flags]

Examples:
  # Activate an Enterprise license (use the currently active deployment).
  $ rasactl enterprise activate

  # Activate an Enterprise license for the 'my-deployment' deployment.
  $ rasactl enterprise activate my-deployment

  # Provide an Enterprise license using STDIN.
  # You can pass an Enterprise license non-interactively by using the --license-stdin flag to provide a license through STDIN.
  # Using STDIN prevents the license from ending up in the shell’s history.
  $ rasactl enterprise activate --license-stdin

Flags:
  -h, --help             help for activate
  -l, --license string   an Enterprise license
      --license-stdin    read an Enterprise license from stdin

The enterprise deactivate command

Deactivate an Enterprise license.

Usage:
  rasactl enterprise deactivate [DEPLOYMENT-NAME] [flags]

Examples:
  # Deactivate an Enterprise license (use the currently active deployment).
  $ rasactl enterprise deactivate

  # Deactivate an Enterprise license for the 'my-deployment' deployment.
  $ rasactl enterprise deactivate my-deployment

Flags:
  -h, --help   help for deactivate

Model Management Commands

You can manage models in Rasa X / Enterprise via rasactl. Below is a list of commands that help with managing models:

$ rasactl help model
manage models for Rasa X / Enterprise

Usage:
  rasactl model [command]

Available Commands:
  delete      delete a model from Rasa X / Enterprise
  download    download a model from Rasa X / Enterprise
  list        list models stored in Rasa X / Enterprise
  tag         tag a model in Rasa X / Enterprise
  upload      upload model to Rasa X / Enterprise

The model delete command

Delete a model from Rasa X / Enterprise.

Usage:
  rasactl model delete [DEPLOYMENT-NAME] MODEL-NAME [flags]

Aliases:
  delete, del

Examples:
  # Delete the 'example-model' model (use the currently active deployment).
  $ rasactl model delete example-model

  # Delete the 'example-model' model for the 'my-deployment' deployment.
  $ rasactl model delete my-deployment example-model

Flags:
  -h, --help   help for delete

The model download command

Download a model from Rasa X / Enterprise to your local machine.

Usage:
  rasactl model download [DEPLOYMENT-NAME] MODEL-NAME [DESTINATION] [flags]

Examples:
  # Download the 'example-model' model (use the currently active deployment).
  # If the destination is not defined, the model will be stored in a current working directory.
  $ rasactl model download example-model

  # Download the 'example-model' model for the 'my-deployment' deployment
  # and store it in the /tmp directory.
  $ rasactl model download my-deployment example-model /tmp/example-model.tar.gz

Flags:
  -h, --help   help for download

The model list command

List all models stored in Rasa X / Enterprise.

Usage:
  rasactl model list [DEPLOYMENT-NAME] [flags]

Aliases:
  list, ls

Examples:
  # List all models (use the currently active deployment).
  $ rasactl model list

  # List all models for the 'my-deployment' deployment.
  $ rasactl model list my-deployment

Flags:
  -h, --help   help for list

The model tag command

Create a tag and assign it to a given model.

Rasa Enterprise allows multiple versions of an assistant to be run simultaneously and served to different users. By default, two environments are defined:

• production
• worker

If you want to activate a model you have to tag it as 'production'.

Learn more here.

Usage:
  rasactl model tag [DEPLOYMENT-NAME] MODEL-NAME TAG [flags]

Examples:
  # Tag the 'my-model' model as 'production' (use the currently active deployment)
  $ rasactl model tag my-model production

  # Tag the 'my-model' with the 'test' tag within the 'my-deployment' deployment.
  $ rasactl model tag my-deployment my-model test

Flags:
  -h, --help   help for tag

The model upload command

Upload a model to Rasa X / Enterprise.

Usage:
  rasactl model upload [DEPLOYMENT-NAME] MODEL-FILE [flags]

Aliases:
  upload, up

Examples:
  # Upload the model.tar.gz model file to Rasa X / Enterprise (use the currently active deployment).
  $ rasactl model upload model.tar.gz

  # Upload the model.tar.gz model file to the 'my-deployment' deployment.
  $ rasactl model upload my-deployment model.tag.gz

Flags:
  -h, --help   help for upload

Upload a model to Rasa X

The following example shows how to download an existing model and upload it via rasactl.

1. Download a model.

$ curl -L https://github.com/RasaHQ/rasa-x-demo/blob/master/models/model.tar.gz?raw=true --output model.tar.gz
[...]

2. Upload the download model to Rasa X.

$ rasactl model upload [deployment name] model.tar.gz

Successfully uploaded.

You can use the rasa model list command to list all available models, e.g

$ rasactl model list [deployment name]
NAME 	VERSION	COMPATIBLE	TAGS	HASH                            	TRAINED AT
model	2.8.2  	true      	none	093dfaad610d330e5f36e6d7dc104d86	05 Aug 21 13:16 UTC

Examples of usage

Run Rasa X / Enterprise with a local Rasa Server

It is possible to run a Rasa X / Enterprise deployment with a local rasa server. The following example shows how to connect a local rasa server that is installed in a Python environment to a running Rasa X / Enterprise deployment.

1. Install rasa on your local machine. More information on how to install rasa you can find in the docs.
2. Activate a Python environment with installed rasa (this step is optional if you don't use a Python environment).

$ source .venv/bin/activate
$ rasa --version
Rasa Version      :         2.7.0
Minimum Compatible Version: 2.6.0
Rasa SDK Version  :         2.8.1
Rasa X Version    :         None
Python Version    :         3.7.11
Operating System  :         Darwin-20.5.0-x86_64-i386-64bit
Python Path       :         /repos/rasa/.venv/bin/python3.7

3. Connect a local rasa server to a Rasa X / Enterprise deployment.

$ rasactl connect rasa
●∙∙ Starting Rasa Server
(production-worker) 2021-08-09 15:56:45 INFO     root  - Starting Rasa server on http://localhost:5005
(production-worker) 2021-08-09 15:56:45 INFO     rasa.model  - Loading model models/20210804-105240.tar.gz...
(production-worker) /Users/tczekajlo/repos/rasa/.venv/lib/python3.7/site-packages/rasa/utils/train_utils.py:565: UserWarning: model_confidence is set to `softmax`. It is recommended to try using `model_confidence=linear_norm` to make it easier to tune fallback thresholds.
  category=UserWarning,
2021-08-09 15:56:56 INFO     rasa.core.brokers.pika  - Connecting to RabbitMQ ...
(production-worker) 2021-08-09 15:56:56 INFO     rasa.core.brokers.pika  - RabbitMQ connection to '127.0.0.1' was established.
(production-worker) 2021-08-09 15:56:56 INFO     root  - Rasa server is up and running.

4. You can check the status of your deployment and see that Rasa version is the same as the rasa version installed locally.

$ rasactl status
Name:                   	funny-hopper
Status:                 	Running
URL:                    	http://funny-hopper.rasactl.localhost
Version:                	0.42.0
Enterprise:             	inactive
Rasa production version:	2.7.0
Rasa worker version:    	2.7.0
Project path:           	not defined

Run Rasa X / Enterprise with mounted a local Rasa project

The example shows how to run Rasa X / Enterprise deployment with mounted a local rasa project.

1. Install rasa on your local machine. More information on how to install rasa you can find in the docs.
2. Create a rasa project

$ rasa init

3. Start a new Rasa X / Enterprise deployment.

$ rasactl start --project

(The rasa start --project command has to be executed in a directory with rasa project. You can use the --project-path flag to pass an absolute path to a rasa project.)

4. Open Rasa X / Enterprise in a web browser.

$ rasactl open

Upgrade Rasa X / Enterprise version

The following example shows how to upgrade Rasa X / Enterprise version for a deployment that already exists.

1. Create the values.yaml file with a specific version.

# values.yaml
rasax:
  tag: "0.42.0"
eventService:
  tag: "0.42.0"
dbMigrationService:
  tag: "0.42.0"

2. Run upgrade.

$ rasactl upgrade deployment-name --values-file values.yaml

Deploy Rasa X in one of the public cloud providers

The following example shows how to deploy Rasa X in one of the major cloud providers. In the example, GCP (Google Cloud Platform) is used.

1. Create a VM using a Linux base image. You can find detailed information on how to create a VM here.
2. Install rasactl on the VM
3. Start a new deployment by executing the rasactl start command.
4. After several minutes you should see details of your deployment.

$ rasactl start
∙∙∙ Ready!

╭ Rasa X ────────────────────────────────╮
│                                        │
│    URL: http://35.184.183.164:30012    │
│    Rasa X version: 0.42.0              │
│    Rasa X password: rasaxlocal         │
│                                        │
╰────────────────────────────────────────╯

Important! The Rasa X / Enterprise deployment will be exposed to the public on one of the service node ports (30000-30100). Remember to add a rule to firewall configuration that allows for access to the Rasa X deployment.

Development

Below you can find a setup required for developing rasactl locally.

How to run it?

1. Install go, e.g. by using brew

$ brew install go

2. Compile it

$ make build

3. Run it

$ ./dist/rasactl

Run unit tests

make test

Kind cluster for developing purposes

1. Install kind and run it

brew install kind

2. Prepare configuration for a kind cluster

$ bash kind/generate-config.sh > config.yaml

3. Create a kind cluster

$ kind create cluster --config config.yaml

After kind is ready, install ingress-nginx:

$ kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/master/deploy/static/provider/kind/deploy.yaml
$ kubectl delete -A ValidatingWebhookConfiguration ingress-nginx-admission

Create a new release

Below you can find steps to prepare a new release.

1. Create a release branch.

git checkout -b prepare_release_<VERSION> main

# e.g. git checkout -b prepare_release_1.5.6 main

2. Update CHANGELOG.md.

VERSION=1.5.6 make release

3. Push the release branch and create a PR

git push origin prepare_release_<VERSION>

4. After the PR is merged create a tag and push it.

git checkout main
git pull
git tag <VERSION> -m "<VERSION>"
git push origin <VERSION>

License

Licensed under the Apache License, Version 2.0. Copyright 2021 Rasa Technologies GmbH. Copy of the license.

rasactl's Issues

Error encountered while uploading a model

Describe the bug
When uploading a trained model from Rasa X UI, I get "Upload failed" with the following error in the logs:

[2022-01-23 08:06:21 +0000] [24] [ERROR] Exception occurred while handling uri: 'http://x.x.x.x/api/projects/default/models'
Traceback (most recent call last):
  File "/usr/local/lib/python3.8/dist-packages/aiohttp/connector.py", line 999, in _create_direct_connection
    hosts = await asyncio.shield(host_resolved)
  File "/usr/local/lib/python3.8/dist-packages/aiohttp/connector.py", line 865, in _resolve_host
    addrs = await self._resolver.resolve(host, port, family=self._family)
  File "/usr/local/lib/python3.8/dist-packages/aiohttp/resolver.py", line 31, in resolve
    infos = await self._loop.getaddrinfo(
  File "uvloop/loop.pyx", line 1469, in getaddrinfo
socket.gaierror: [Errno -2] Name or service not known

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/usr/local/lib/python3.8/dist-packages/sanic/app.py", line 973, in handle_request
    response = await response
  File "/usr/local/lib/python3.8/dist-packages/rasax/community/api/decorators.py", line 247, in decorated_function
    return await await_and_return_response(args, kwargs, request)
  File "/usr/local/lib/python3.8/dist-packages/rasax/community/api/decorators.py", line 177, in await_and_return_response
    response = await response
  File "/usr/local/lib/python3.8/dist-packages/rasax/community/api/decorators.py", line 84, in decorated_function
    return await f(request, user=user, *args, **kwargs)
  File "/usr/local/lib/python3.8/dist-packages/rasax/community/api/blueprints/models.py", line 69, in upload_model
    minimum_version = await model_service.minimum_compatible_version()
  File "/usr/local/lib/python3.8/dist-packages/rasax/community/services/model_service.py", line 182, in minimum_compatible_version
    info = await stack_service.version()
  File "/usr/local/lib/python3.8/dist-packages/rasax/community/services/stack_service.py", line 124, in version
    response = await session.get(
  File "/usr/local/lib/python3.8/dist-packages/aiohttp/client.py", line 520, in _request
    conn = await self._connector.connect(
  File "/usr/local/lib/python3.8/dist-packages/aiohttp/connector.py", line 535, in connect
    proto = await self._create_connection(req, traces, timeout)
  File "/usr/local/lib/python3.8/dist-packages/aiohttp/connector.py", line 892, in _create_connection
    _, proto = await self._create_direct_connection(req, traces, timeout)
  File "/usr/local/lib/python3.8/dist-packages/aiohttp/connector.py", line 1011, in _create_direct_connection
    raise ClientConnectorError(req.connection_key, exc) from exc
aiohttp.client_exceptions.ClientConnectorError: Cannot connect to host gateway.docker.internal:5005 ssl:default [Name or service not known]

Uploading a model with rasactl model upload also fails with the following error:

Error: rasa server is not connected to the production environment

To Reproduce
Steps to reproduce the behavior:

1. Install Rasa X following the docs here.
2. Install rasa with pip install rasa.
3. Create new project with rasa init and cd to the project.
4. Upload the model from the UI or using rasactl model upload

Desktop (please complete the following information):

• OS: Ubuntu 20.04

rasactl version*
rasactl version 1.0.1

rasactl open fails when there is no browser

Describe the bug
rasactl open fails when there is no browser

$ rasactl open
Error: Can't open the URL http://127.0.0.1:4321 in your web browser: exec: "xdg-open,x-www-browser,www-browser": executable file not found in $PATH
...

**Expected behaviour **

$ rasactl open
http://127.0.0.1:4321
...

rasactl version

• rasactl --version

feature flag --kubeconfig-context

In case the kubeconfig has multiple clusters configured - rasactl selects the last added cluster.

which can lead to false environment deployments.

would be good to introduce a context selector and a check with warning when multiple clusters are configured and none is selected to improve the overall experience and to lower the risk of false deployments.

`~/.kube/config` file is missing

I just got this error after following the installation instructions in the init branch.

(venv) ➜  rasactl git:(init) ✗ sudo ./rasactl start -p
Error: stat /root/.kube/config: no such file or directory
stat /root/.kube/config: no such file or directory

It seems like for local mode I need a ~/.kube/config file around? If that's the case, it would be good to add a section in the installation guide on how to add it. I imagine this is a file that's known for kubernetes users, but since kubernetes novices will also need to use this tool I'd argue we need to add a step to the installation guide.

[CI] Add dependabot

• Add configuration for dependabot

How to access Rasa X web interface?

Hi, I am just deploy my Rasa X server on my server. It give me a random URL like example below.

$ rasactl status vibrant-yalow
Name:                   	vibrant-yalow
Status:                 	Running
URL:                    	http://vibrant-yalow.rasactl.localhost
Version:                	0.42.0
Enterprise:             	inactive
Rasa production version:	2.8.1
Rasa worker version:    	2.8.1
Project path:           	/home/ubuntu/test

But I can not access it by URL: http://vibrant-yalow.rasactl.localhost. How can I

Thank you so much and have a great day!

Add a `--background` option to `rasactl connect rasa`

rasactl connect rasa starts a rasa server in the foreground, so if you close the terminal, your server dies.
Would be helpful to have a --background flag so rasactl connect rasa --background starts a rasa server in the background independent of terminal session.

Raise a warning on Docker CLI <= 20.10.0

What problem are you trying to solve?

rasactl ultilizes host-gateway, which is a feature added since docker cli 20.10.0.

When users have Docker CLI <= 20.0.3, they will see Error response from daemon: invalid IP address in add-host: "host-gateway".

What's your suggested solution?

Throw a warning on Docker CLI <= 20.10.0

Examples (if relevant)

No response

Is anything blocking this from being implemented? (if relevant)

No

Definition of Done

No response

[`connect rasa`] Add support for action endpoint

What problem are you trying to solve?

Currently, the connect rasa doesn't support Rasa Action Server.

What's your suggested solution?

No response

Examples (if relevant)

No response

Is anything blocking this from being implemented? (if relevant)

No response

Definition of Done

The connect rasa supports Action Server.

rasactl should use the latest rasa-x-helm by default

What problem are you trying to solve?

In rasactl 1.0.2, rasactl start -h shows:

at the time of writing, the latest rasa-x helm chart is 4.3.2.

rasactl should automatically use the latest rasa-x helm chart so that customers are able to use the latest features and not get stuck troubleshooting a bug that was already fixed in a subsequent release.

What's your suggested solution?

rasactl should always use the latest rasa-x helm release

Examples (if relevant)

No response

Is anything blocking this from being implemented? (if relevant)

No response

Definition of Done

rasactl uses the latest rasa-x helm release by default

Emit notification if a newer version is available.

What problem are you trying to solve?

Print a notification that informs a newer version is available.

What's your suggested solution?

No response

Examples (if relevant)

No response

Is anything blocking this from being implemented? (if relevant)

No response

Definition of Done

No response

Model training lasts forever over Rasa X UI

Describe the bug
If you connect a Rasa OSS to Rasa X using the connect rasa command, model training triggered via Rasa UI X lasts forever. It is caused by an incorrect URL that is used by Rasa OSS to send results back.

Pip installable

I just installed go in order to use this tool and it made me wonder ... is there a good reason why rasactl is written in go? I'm mainly asking because it'll be a hurdle for the average python developer. Is it possible to turn this CLI into something that is also pip install-able?

Deployment fails if the default namespace already contains helm release

If the default namespace already contains the rasa-x release, then deployment fails with another operation (install/upgrade/rollback) is in progress even if the target namespace is different.

Add additional deployment statuses

What problem are you trying to solve?

Currently, if Rasa X is deploying or upgrading a deployment, rasactl status/list returns status for a given deployment as Stopped.

What's your suggested solution?

Add separate statuses for installing deployment and upgrading.

Examples (if relevant)

No response

Is anything blocking this from being implemented? (if relevant)

No response

Definition of Done

No response

Increase unit tests coverage

Increase unit tests coverage.

Cannot connect to a local rasa open source server

Describe the bug

I get the following error message when running rasactl connect rasa.

panic: interface conversion: interface {} is nil, not map[string]interface {}

goroutine 1 [running]:
github.com/RasaHQ/rasactl/pkg/k8s.(*Kubernetes).GetRabbitMqCreds(0xc0004a6500)
        /go/src/rasactl/pkg/k8s/secrets.go:127 +0x23a
github.com/RasaHQ/rasactl/pkg/rasactl.(*RasaCtl).saveRasaEndpointsFile(0xc0004a6280, {0xc0010fe840, 0x20})
        /go/src/rasactl/pkg/rasactl/connect_rasa.go:280 +0x12b
github.com/RasaHQ/rasactl/pkg/rasactl.(*RasaCtl).ConnectRasa(0xc0004a6280)
        /go/src/rasactl/pkg/rasactl/connect_rasa.go:142 +0xdc9
github.com/RasaHQ/rasactl/cmd.connectRasaCmd.func2(0xc00029bb80, {0xc0000b8d20, 0x1, 0x3})
        /go/src/rasactl/cmd/connect_rasa.go:117 +0xab
github.com/spf13/cobra.(*Command).execute(0xc00029bb80, {0xc0000b8c90, 0x3, 0x3})
        /go/pkg/mod/github.com/spf13/[email protected]/command.go:856 +0x60e
github.com/spf13/cobra.(*Command).ExecuteC(0x3eff7e0)
        /go/pkg/mod/github.com/spf13/[email protected]/command.go:974 +0x3bc
github.com/spf13/cobra.(*Command).Execute(...)
        /go/pkg/mod/github.com/spf13/[email protected]/command.go:902
github.com/RasaHQ/rasactl/cmd.Execute()
        /go/src/rasactl/cmd/root.go:75 +0x25
main.main()
        /go/src/rasactl/main.go:32 +0xc5

To Reproduce
Steps to reproduce the behavior:

1. Install Rasa X following the docs here.
2. Install rasa with pip install rasa.
3. Create new project with rasa init and cd to the project.
4. Run rasactl connect rasa and see the output error.

Expected behavior
Rasa X should connect to Rasa OSS successfully.

Desktop (please complete the following information):

• OS: VM with Ubuntu 20.04 (4vCPU and 16GB RAM).

rasactl version*
rasactl version 1.0.0