Hi, I think I’ve seen this answered before, but I was wondering if rasa was multithreaded. If there were many many requests per second, how well would it perform? And what’s the best way for scaling? Is the recommended method to dockerize and scale that way?


As for handling multiple requests per second, this is handled out of the box.

Yes the recommended way is to dockerize, this is how we handle our production deployment setup for the Rasa Platform.

What are the system requirements for the platform system for example to handle 100 concurrent users, 1000 concurrent users

The nginx container of the platform is setup to handle that amount of concurrent users. The actual system requirements of the platform shouldn’t differ too much from what we’ve specified in our documentation.