That "100k hits" figure you quoted surprised me, so after going over their page, I just wanted to add here that that's "100K hits/day", which translates to just

That "100k hits" figure you quoted surprised me, so after going over their page, I just wanted to add here that that's "100K hits/day", which translates to just 1 query per second, which is unreasonably conservative. In a post [0] from last year, Expensify tell how (with some modifications) they managed to get sqlite to 4M qps on a single powerful server running 192 cores. That's 20K qps/thread for their huge table, and they got over 200K qps/thread for a table with a single row (demonstrating how little overhead sqlite has). [0] https://blog.expensify.com/2018/01/08/scaling-sqlite-to-4m-qps-on-a-single-server/ https://blog.expensify.com/2018/01/08/scaling-sqlite-to-4m-q...