Hacker News new | ask | show | jobs
by sgarland 246 days ago
Many distributed databases have mechanisms to use an auto-increment, actually - often, generating large chunks at a time to hand out.
1 comments

Our “distributed database” at a fortune 90 company spans at least 10 different database products.

UUIDv4 lets us sidestep this.

Is it bad design? Probably. Is it going to happen at huge companies? Yes.

You’re not wrong, of course. It’s a natural consequence of the eschewing of DBAs, and the increasingly powerful compute available - even if someone did notice that the slowdown was due to the PK choice, they can often “fix” that by paying more money.