Re: Are tables independent?

[Rick James]

> For example, imagine SELECT from a table with one row. Is this query different (its speed, resource usage, etc)

Each table is effectively independent of all other tables. You can even see the tables in the file system. (InnoDB without file_per_table puts all tables in a single tablespace, but still the "blocks" for each table are independent.)

SELECTing one using an appropriate INDEX will be nearly as fast whether the table has one row or one billion rows. Most indexes are BTrees; a tiny table will have 1 level of BTree; a billion row table will have about 5 levels. Hence, it is not completely as fast.

A single SELECT can JOIN or UNION multiple tables.

Almost always a table should have index(es).

Bottom line: design your tables/databases/etc in a way that makes sense for your application. If you hit any limits (unlikely) or have performance problems, we can discuss remedies.