Hey guys, Im developing two different chatbots and I have them on a Google Platform VM. I want to know how to access to them like this: -> https://mydomain.com/chatbot1 -> https://mydomain.com/chatbot2 Instead of having: -> https://mydomain.com:5001 -> https://mydomain.com:5002 I want to do that so I can run multiple chatbots on the same VM and I dont expose the ports. After doing this the idea is to insert the chatbot on a 3rd user web like this: <script> WebChat.default.init({ selector: "#webchat", initPayload: "XXX", interval: 1000, customData: {"userId": "123"}, socketUrl: "http://mydomain.com/chatbot1", socketPath: "/socket.io/", title: "XXX", subtitle: "XXX", inputTextFieldHint: "XXX", connectingText: "XXX", hideWhenNotConnected: true, fullScreenMode: false, showFullScreenButton: false, profileAvatar: "xxx.jpg", params: { images: { dims: { width: 250, height: 200, } }, storage: "XXX", }, }) </script> Both need to run through HTTPS. Is this possible? Do I need Apache or something similar? If this is the case, how can I configure it? I dont know if it is relevant but I only have access to the VM through SSH.

I suggest you read more about nginx . You’ll need a config file for nginx, you can use something similar to the following: server { listen 443 default_server; server_name mydomain.com; ssl on; ssl_certificate /path/to/certificate; ssl_certificate_key /path/to/certificate/key; ssl_session_cache shared:SSL:10m; location /chatbot1/ { proxy_pass http://localhost:5001/; proxy_http_version 1.1; proxy_set_header Upgrade $http_upgrade; proxy_set_header Connection "upgrade"; proxy_set_header Host $host; } location /chatbot2/ { proxy_pass http://localhost:5002/; proxy_http_version 1.1; proxy_set_header Upgrade $http_upgrade; proxy_set_header Connection "upgrade"; proxy_set_header Host $host; } } server { listen 80 default_server; server_name mydomain.com; return 301 https://$host$request_uri; } Note: I see that you want to use the socketio channel, hence the config of nginx above has been modified to upgrade the http connections to websocket connection.

Thanks! Could you give me an example of this with the code that I put above?

Thank you very much! I tried it. Now I have nginx running on my server with the code that you gave me. But when I change from this: socketUrl: "https://chatbot.example.com.ar:5004", to this: socketUrl: "https://chatbot.example.com.ar/chatbot1", on my webpage, it doesnt work. My whole nginx.conf file looks like this: user ib; worker_processes auto; error_log logs/error.log; pid logs/nginx.pid; include /etc/nginx/modules-enabled/*.conf; events { worker_connections 768; # multi_accept on; } http { sendfile on; tcp_nopush on; tcp_nodelay on; keepalive_timeout 65; types_hash_max_size 2048; include /etc/nginx/mime.types; default_type application/octet-stream; ssl_protocols TLSv1 TLSv1.1 TLSv1.2; # Dropping SSLv3, ref: POODLE ssl_prefer_server_ciphers on; access_log /var/log/nginx/access.log; error_log /var/log/nginx/error.log; gzip on; include /etc/nginx/conf.d/*.conf; include /etc/nginx/sites-enabled/*; server { listen 443 ssl default_server; server_name chatbot.example.com.ar; ssl on; ssl_certificate /etc/letsencrypt/live/chatbot.example.com.ar/fullchain.pem; ssl_certificate_key /etc/letsencrypt/live/chatbot.example.com.ar/privkey.pem; ssl_trusted_certificate /etc/letsencrypt/live/chatbot.example.com.ar/chain.pem; ssl_session_cache shared:SSL:10m; location /chatbot1/ { proxy_pass http://localhost:5004/; proxy_http_version 1.1; proxy_set_header Upgrade $http_upgrade; proxy_set_header Connection "upgrade"; proxy_set_header Host $host; } location /chatbot2/ { proxy_pass http://localhost:5006/; proxy_http_version 1.1; proxy_set_header Upgrade $http_upgrade; proxy_set_header Connection "upgrade"; proxy_set_header Host $host; } } server { listen 80; server_name chatbot.example.com.ar; return 301 https://$host$request_uri; } } I dont know if I am doing something wrong or what. Maybe I should add the map and upstream websocket as its shown here ? Please if you can help me I would be very glad :grin::grin: Cheers, Facu

Could you change your script to this and try again? <script> WebChat.default.init({ selector: "#webchat", initPayload: "XXX", interval: 1000, customData: {"userId": "123"}, socketUrl: "https://mydomain.com", socketPath: "/chatbot1/socket.io/", title: "XXX", subtitle: "XXX", inputTextFieldHint: "XXX", connectingText: "XXX", hideWhenNotConnected: true, fullScreenMode: false, showFullScreenButton: false, profileAvatar: "xxx.jpg", params: { images: { dims: { width: 250, height: 200, } }, storage: "XXX", }, }) </script>

I have just tried it but it seems that doesn’t work neither. Just to be sure, I run this two commands: rasa run actions rasa run --ssl-certificate /xxx/fullchain.pem --ssl-keyfile /xxx/privkey.pem --port 5004 -vv --cors '*' --enable-api --model xxx --endpoints /xxx/endpoints.yml --credentials /xxx/credentials.yml And doing this it works if I put this: socketUrl: "https://chatbot.example.com.ar:5004", So I dont think that the problem is a port misunderstanding :thinking:

Well I remember doing this sometime ago and it worked for me. I’ll check day after tomorrow and see if I’m missing anything.

Im going to try without the ssl arguments in rasa run and I will tell you. I really appreciate your time and answers, they are helping me a lot. Thank you very much!!

You are welcome. I see that you have declared your server in nginx.conf file. I think the correct way to do this is to create a seperate conf file inside the sites-available directory and create a symbolic link to it in the sites-enabled directory. Atleast that’s how I did it back then :thinking:

Hi! Finally it works!! Thank you very much. This wouldn’t be possible without your help. :hugs::hugs: I’m going to list the things I have done to sum up for future users that might have the same question: Install nginx In your /etc/nginx/sites-available/default add the code that @saurabh-m523 put above ( make sure that you have a symbolic link to this file in /etc/nginx/sites-enabled/) On your html file, you should put this: socketUrl: "https://mydomain.com/", socketPath: "/chatbot1/socket.io/", And finally, you should run rasa without ssl certification Thank you ver much @saurabh-m523 !!!

You are welcome! :slightly_smiling_face:

Running multiple chatbots on same server and hiding ports

Rasa Open Source

saurabh-m523 (Saurabh Kumar) March 9, 2020, 4:56pm 8

You don’t need to run rasa with ssl arguments, that’s why we are configuring nginx to use https and act as a reverse proxy to rasa server.

Topic		Replies	Views
Deploy rasa on a ssl enabled server Rasa Open Source	18	3764	August 1, 2022
Run multiples chatbot in the same server and connect through HTTPS Rasa Open Source	2	492	March 6, 2020
Facing a SSL error while using Rasa-Webchat on https websites Rasa Open Source	26	1796	October 28, 2021
Multiple chatbots on a server with docker [Deprecated] Rasa X Community Edition	1	725	October 9, 2020
Rasa docker nginx revers proxy webchat issue Rasa Open Source	0	510	March 19, 2021

Running multiple chatbots on same server and hiding ports

Related topics