Is lemmy scalable?

Preface: I’m new to the project and still trying to install my first server, and the instructions so far are leading into problems. I’m a professional Technical Editor since the early 1990’s, so I’m trying to document the problems over on https://lemmy.ml/post/1160483 about a from-scratch Ubuntu 22.04.2 LTS install.

Based on what I have seen, I have big concerns about scale. Using a SQL database for a messaging system like this with all these granular real-time data elements like vote counts and dynamic sorting orders I think is going to prove to have serous scaling problems without some concept of caching results in a simple format (like a plain old disk file or NoSQL database) before returning results to the webapp clients or API. I am struggling with my human languages lately, but I did start to raise some topics over on community [email protected]

I understand the desire for a real-time database that has exact up to date information on every vote count and every single latest comment, but I think server operators are going to run into big problems with Lemmy code as it works now once their data starts to get into the thousands of postings per day, tens of thousands of user profiles being looked up for render those postings, and millions of comments. When there isn’t enough RAM for those SQL index keys, things will dramatically change in performance.