I'm familiar with selecting random rows from a table in PostgreSQL. It's really slow. I've looked into some other databases and they all seem to be bad at it. Mostly because it's not a normal use case for any of them.
Is there any type of database that can select a random row quickly? The only hard requirement is that it be installable on Linux.
So far I have a table with 500 million rows and anticipate it growing to a few billion in the near future and would like to be able to select a random item. I'd also like to be able to delete the same row a short time later, so it'll need some sort of key/index for that.
Right now my psql schema has 3 columns: a text field, a date, and an integer sequence id. The date is not important and doesn't need to be in the new database.