Rasa 2.0 Custom Connector deployment on Rasax (33.x) docker

[Deprecated] Rasa X Community Edition
1

I am using a custom connector that works when I run rasa locally, however, when I try to deploy the custom connector on Rasax (33.x) I am receiving and error:

RasaException: Failed to find input channel class for 'connector.custom_connector.CustomConnector'. Unknown input channel. Check your credentials configuration to make sure the mentioned channel is not misspelled. If you are creating your own channel, make sure it is a proper name of a class in a module.

I installed Rasax on my machine at /etc/rasa/ using: curl -sSL -o install.sh https://storage.googleapis.com/rasa-x-releases/0.33.1/install.sh

I created a dir connector/ where I placed my custom_connectory.py file.

I also placed the following line in credentials.yml: connector.custom_connector.CustomConnector:

What am I missing?

2

Even I am stuck on the same issue since days.

Issue similar to this. (the one I am facing)

@erohmensing kindly see if you can help or redirect someone who can.

3

@nmvl were you able to find a solution to this?

4

No luck yet :confused:

5

@nvml did you mount the directory as a volume?

6

Kindly elaborate on how to do that. I religiously followed the steps specified by @Arjaan here but didn’t come across any command mounting directory as volume.

7

let’s discuss your issue in that thread.

8

Hi Akela, I am not sure what you mean. Do you have a set of instructions that you are referring to that you could point me at? I do not have a Dockerfile if that is what you are referring to.

These are the files that I have after running the install.sh script I mentioned in my initial post.

rasax_files
rasax_files1555×20 5.95 KB

I added my custom_connector.py and __init__.py in the connector directory, then ran docker-compose up -d.

And added the line as mentioned in my initial post to the credentials.yml file.

Does that help?

9

Any update on this?

10

Sorry @nmvl this slipped through the cracks. You need to mount this directory as a volume to the image, otherwise the image won’t know about that file. You can read more about volumes here: Use volumes | Docker Documentation

11

could you please help me out a little more?

12

A easy way to avoid all qustion is build yourself server (like handle channel),and use rasa as a nlu server.Believe me ,u won’t feel better more.
No rasa x,no channel, no custom action.My team hava used this way to server millions of user.
this suggestion may not help your issue,but can help u save more time in the future