Connecting Telegram using Rasa X and Google Cloud server

Hi there, I have followed the Udemy workshop for advanced deployment workshop: Then I have hosted Rasa X in a Google cloud and I have deployed Rasa X using helm chart with values.yml file containing the following line for rasa and rasa X :

  • Rasa X : 0.38.1
  • Rasa : 2.4.0

I have connected my git, train a model in Rasa X and it works perfectly. But instead of sharing my external IP address with my Google VM I would like to share my bot with Telegram. I have found that I have to add external Channels in values.yml in my VM instead of credentials.yml in my git. Then I have tried this in my values.yml:

rasa:
    # token Rasa accepts as authentication token from other Rasa services
    token: "..."
    # tag refers to the Rasa image tag
    tag: "2.4.0-full"
    additionalChannelCredentials:
        telegram:
            access_token: "..."
            verify: "..."
            webhook_url: "http://<Rasa X external IP Address>/webhooks/telegram/webhook"

But it doesn’t work, has anybody found a tutorial for this case? I already found tutorial for local Rasa X ( Custom channels when using the RasaX Helm Chart) or using ngrock (in rasa masterclass youtube video) but I would like to just link my Rasa X to an external channel after the Udemy formation.

Thank you for your help, tell me if you need more information but i have strictly followed the Udemy formation above

See you

Julien

Hi Julien! First of all, which endpoint are you trying to send messages to where it is not working?

That does look like the correct way to provide additional channel credentials. To clarify, did you also run a helm upgrade after updating your channels?

In debug mode, the rasa-production server will list all of its available endpoints upon startup. If you add the debugMode: true value to your values.yml and redeploy, that is a good way to check whether the telegram channel was registered or not.