I got my 5 rasa chatbots running in different docker containers but when I try use webchat it fails. I can run rasa with webchat in normal linux server.
I don’t get error messages, just no webchat chat window.
Here is command I run
docker run --name=rasa1 --user 1000 -v $(pwd):/app rasa/rasa:1.10.11-full run -m models --enable-api --cors “*” --debug
Here is my environment
Ubuntu 18.04
Docker 19.03.12
and I also added socketio to credentials.yml as @InnoOmnia did.
I am getting this Error in app_1 container:
INFO:sanic.root:Goin' Fast @ http://0.0.0.0:5055
INFO:sanic.root:Starting worker [1]
ERROR:sanic.root:Exception occurred while handling uri:
'http://172.24.0.9:5055/socket.io/?EIO=3&transport=polling&t=NHHz8Jc'
Traceback (most recent call last):
File "/opt/venv/lib/python3.7/site-packages/sanic/app.py", line 940, in handle_request
handler, args, kwargs, uri, name = self.router.get(request)
File "/opt/venv/lib/python3.7/site-packages/sanic/router.py", line 411, in get
return self._get(request.path, request.method, "")
File "/opt/venv/lib/python3.7/site-packages/sanic/router.py", line 475, in _get
raise NotFound("Requested URL {} not found".format(url))
sanic.exceptions.NotFound: Requested URL /socket.io/ not found
and
rasa_rasa-worker_1 exited with code 132
rasa_rasa-production_1 exited with code 132
How can I pass those params to the docker-compose to the specific container?
-m models --enable-api --cors “*” --debug
I am a little lost, don’t know where to start to tackle this issue.
Should I ask another question? Related to the error.
I’m using Rasa open source, not Rasa X and if you are using Rasa X then you don’t need that command chain. If you use Rasa open source then why you need docker-compose then?
Could you elaborate on this statement? If your question was on how to pass arguments to rasa server from Docker, then use the below command in your Dockerfile
SocketIO doesn’t need any params to be specified to the rasa server. Just provide the config in credentials.yml file and you can connect using the /socket.io endpoint.
I’m not using Rasa-X. I wrote a docker-compose my self for rasa server, actions, nginx, and a node app for webchat. The image I’ve shared is from that setup.
I have one server setup with Rasa-X and its working properly.
These are the steps I’ve used. Could you try them?
sudo apt update && sudo apt upgrade
curl -sSL -o install.sh https://storage.googleapis.com/rasa-x-releases/0.32.0/install.sh
sudo bash ./install.sh
cd /etc/rasa
sudo docker-compose up -d
sudo python3 rasa_x_commands.py create --update admin me <ADMIN_PASSWORD>
For Rasa Actions Server execute the following steps from /etc/rasa
sudo mkdir actions
sudo touch actions/__init__.py
sudo nano actions/actions.py
COPY actions.py code to this file
sudo nano docker-compose.override.yml
It could have something to do with the CPU and AVX Instructions.The default Tensorflow build might not be able to run on old CPUs that does not support AVX instructions. See docs.