Load when it is necessary to leave sqlite?

Question

SQLite

At what load on the site is it worth switching from sqlite to mysql/postgre?

Plot: There is a static site (nginx), hanging on a minimal virtual machine on DO. The site is generated by several spider demons on a schedule. The customer wanted to give these spiders a task remotely and view / delete the results of the execution. This is approximately 300/3000 distributed (several users not connected to each other) write / read requests per day. There was a need to use flask + database. Because The virtual machine is weak in memory, and even put mysql there, so the question arises: is it worth switching to a more productive virtual machine or using sqlite?

Answer 1

SQLITE can only have problems with writing. There is a special log that can slow down. But it can be turned off. There are probably a lot more settings.
You can also use write queues, if applicable.
I made traffic accounting systems, where 10-20 thousand clicks per day (from teaser networks) were recorded in the database normally. True, this experiment lasted only a week.
Another experience with a lower load (up to several thousand transitions a day) is in its third year with over a million entries.
I don't know how useful this will be for you.
And yes, SQLITE is a full-fledged DBMS, only embedded, not client-server. Although there are servers for SQLITE, and you can write your own server for a specific task, if you feel like it.

Answer 2

Without knowing your task and the distribution of requests over time, it is impossible to give an unambiguous answer. Do a load test.
From general considerations, I can say that the transition to a full-fledged database is usually more useful.