How to design a database architecture for a simple SaaS solution?

Question

Python

Given:
Not a complex application (Flask + gevent + uwsgi + nginx + front in JS) that is conceived as a service (something like a logger with additional features)
SQLite is currently used as a database, but the architecture provides for switching to almost any SQL base.
This is not a social network, so there will be no over 9000 clients. The database is, to put it mildly, simple. About 20 tables.
There is a proposal for each client to start a separate database. Here are the arguments:
1. So that more active clients do not interfere with less active ones. Using all sorts of load balancers, too, I think, is not our scale
.
3. In such a model, SQLite will be quite enough, which, as you know, does not require additional resources for the database server. That is, the service can be deployed on simpler / cheaper hardware.
Question:
Store data in one or several databases?
I have no experience in creating Saas solutions ... Perhaps my arguments are "far-fetched" ...

Answer 1

no separate, 100 clients - 100 connections - how do you imagine that?
well, SQLIte multiconnect does not tolerate well, you need a classic - Muskul or Postgre
or Mongo