Rasa-x 1.0.1 compatibility issue with Rasa 2.8.2 & 2.8.17 models

I am using Ubuntu 20 and following the most recent install guidelines for v1.0.0 and the Ephermeral InstallerIntroduction .

╰─❯ cat values.yml                                                                                                                                                                                                                       ─╯
rasax:
  tag: "1.0.1"
eventService:
  tag: "1.0.1"
dbMigrationService:
  tag: "1.0.1"

╰─❯ rasactl status                                                                                                                                                                                                                       ─╯
Name:                           rasa-x-ben                         
Status:                         Running                            
URL:                            http://rasa-x-ben.rasactl.localhost
Version:                        1.0.1                              
Enterprise:                     inactive                           
Rasa production version:        0.0.0                              
Rasa worker version:            0.0.0                              
Project path:                   /data/Software/rasa-x-1.0.0   
rasactl start --values-file values.yml --kubeconfig /home/ben/.kube/config --project rasa-x-ben --verbose

I have created a Rasa Open Source model using

╰─❯ pip freeze | grep rasa                                                                                                                                                                                                               ─╯
rasa==2.8.2
rasa-sdk==2.8.3

and rasa init

I can log into the GUI but when I upload the model from file it says " Upload failed

Something went wrong. Please try again." and the logs for the Rasa X server shows:

[2021-12-22 09:56:21 +0000] [27] [ERROR] Exception occurred while handling uri: 'http://rasa-x-ben.rasactl.localhost/api/projects/default/models'
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()
AttributeError: 'NoneType' object has no attribute 'version'
[2021-12-22 09:56:21 +0000] - (sanic.access)[INFO][10.244.0.7:56126]: POST http://rasa-x-ben.rasactl.localhost/api/projects/default/models  500 213

It sounds like a compatibility error, but as far as I can see these versions of Rasa Open Source and Rasa-X should be compatible? Compatibility Matrix

Many thanks for any help on this

@holyone2two try update the rasa and rasa-sdk with upper version please but less then rasa 3.x as rasa x is not supporting the latest version yet ? Hope that will solve the issue.

Thank you for this suggestion, but I have read that Rasa and Rasa-SDK 3+ is not currently compatible with Rasa-X at all, so I purposely am not using the latest Rasa version (3.0.0).

This is suggested here and on the forums.

Is the documentation wrong?

@holyone2two yeps you are right, my bad I am still on rasa 2.X that why said that update to latest, basically I mean with upgrade version and less then 3.x. Let me update above.

@holyone2two yes I am sure it related to rasa and rasa-sdk version my combination is 2.8.1 and 2.8.2 with rasa x 0.42.4

I should note that I spotted there are later versions of Rasa such as pip install rasa==2.8.17 - but when I try this upgrade, the pip dependency resolver gets stuck

@holyone2two can you share some error screenshot please ?

I managed to update my Rasa Open Source version to latest minor fix subversion 2.8.17:

╰─❯ pip freeze | grep rasa                                                                                                                                                                                                               ─╯
rasa==2.8.17
rasa-sdk==2.8.3

By creating a fresh python 3.8 conda environment, and upgrading pip, before installing it.

Even when I freshly generate (rasa train) the model file with this, I get the same error though.

[2021-12-23 18:05:49 +0000] [25] [ERROR] Exception occurred while handling uri: 'http://rasa-x-ben.rasactl.localhost/api/projects/default/models?limit=50&offset=0'
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/blueprints/models.py", line 44, in get_models
    models, total_models = await _model_service(request).get_models(
  File "/usr/local/lib/python3.8/dist-packages/rasax/community/services/model_service.py", line 824, in get_models
    minimum_compatible_version = await self.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()
AttributeError: 'NoneType' object has no attribute 'version'
[2021-12-23 18:05:49 +0000] - (sanic.access)[INFO][10.244.0.7:49044]: GET http://rasa-x-ben.rasactl.localhost/api/projects/default/models?limit=50&offset=0  500 213
[2021-12-23 18:05:49 +0000] - (sanic.access)[INFO][10.244.0.7:49042]: GET http://rasa-x-ben.rasactl.localhost/api/health  200 144
[2021-12-23 18:05:58 +0000] - (sanic.access)[INFO][10.244.5.1:41486]: GET http://10.244.5.3:5002/  200 3322
[2021-12-23 18:05:58 +0000] - (sanic.access)[INFO][10.244.5.1:41488]: GET http://10.244.5.3:5002/  200 3322
[2021-12-23 18:06:07 +0000] [22] [ERROR] Exception occurred while handling uri: 'http://rasa-x-ben.rasactl.localhost/api/projects/default/models'
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()
AttributeError: 'NoneType' object has no attribute 'version'
[2021-12-23 18:06:07 +0000] - (sanic.access)[INFO][10.244.0.7:49272]: POST http://rasa-x-ben.rasactl.localhost/api/projects/default/models  500 213

Any ideas much appreciated, as I am baffled how to use Rasa X 1.0.1

I also tried a model trained with Rasa open source v3 and that had same issue.

@holyone2two Right, rasa v3 is not Compatible with Rasa X, and from the above traceback - please see the compatibility issue and fix that I guess that is the main issue and train the model.

@holyone2two confirm me that you are working on local machine is that right?

I am aware that v3 is not compatible and that’s why I used v2, (see the pip freeze in my posts above for specific versions) but given the error is seemingly version related I just made a note that I’ve tried v3 too.

Yes I’m testing on local Ubuntu 20 machine with Conda environment for python

@holyone2two what’s your issue then?

When using Rasa Open Source 2.8.17 with Rasa X 1.01 which are compatible according to the compatibility chart, I get the error pasted here

@holyone2two you can try downgrade rasa to 2.8.14?

I have tried downgrading to 2.8.14

╰─❯ pip freeze | grep rasa                                                                                                                                                                                                                                                  ─╯
rasa==2.8.14
rasa-sdk==2.8.3

then doing a rasa train operation to get the new model file. But get the same error

[2021-12-25 16:48:39 +0000] [22] [ERROR] Exception occurred while handling uri: 'http://rasa-x-ben-2.rasactl.localhost/api/projects/default/models'
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()
AttributeError: 'NoneType' object has no attribute 'version'

After connecting Rasa-X to github version control, and clicking add model > train model, I also get the following error:

[2021-12-25 17:01:33 +0000] [21] [ERROR] Exception occurred while handling uri: 'http://rasa-x-ben-2.rasactl.localhost/api/projects/default/models/jobs'
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 249, in train_model
    stack_service = stack_services[environment]
KeyError: 'worker'

I saw you had it working with these versions, so I downgraded my rasa-X to 0.42.4, and my Rasa Open Source to 2.8.2 and started a new Rasa X server, but uploading a freshly trained model gives the same error as before:

AttributeError: 'NoneType' object has no attribute 'version'
'NoneType' object has no attribute 'version

My best guess is that Rasa-X is expecting some configuration that I am not providing.Should my values.yml provide anything beyond simply the version of Rasa-X I wish to run?

Does anyone have any advise on this? Spent days trying to run Rasa-X with Rasa Open Source, and given I am following the official documentation to the letter, I find it concerning I cannot get it to run.

Hi Benjamin @holyone2two, I have a very similar error. I am installing RASA-X on a VM on digitalocean (ubuntu 20.4). I used the new Rasa Ephemeral Installer (REI) to install RASA-X without any custom configuration. When training models I get the same NoneType object has not attribute version and KeyError: worker. I asked my question here: Rasa X Unauthorized when training model

Did you find any solution? I am stuck for now :confused:

I am afraid I have no solution yet. This seems like a fundamental bug in the current codebase to me, given it makes Rasa-X unusable and I have followed official docs to install. There has been no support for this issue from Rasa developers on this over the last few weeks it would seem.

I am having the same issue after deploying Rasa X with Helm. I hope that a solution comes really soon.

1 Like