Вы можете использовать оконные функции для этого и мод арифметики.Для простого повторения:
with a as (
select a.*, rownum as seqnum
from a
),
b as (
select b.*, rownum as seqnum, count(*) over () as cnt
from b
)
select a.col, b.col
from a join
b
on mod(a.seqnum - 1, b.cnt) = b.seqnum - 1;
Для более случайного назначения, рандомизируйте следующие числа:
with a as (
select a.*,
row_number() over (order by dbms_random.random) as seqnum
from a
order by db
),
b as (
select b.*, count(*) over () as cnt,
row_number() over (order by dbms_random.random) as seqnum
from b
)
select a.col, b.col
from a join
b
on mod(a.seqnum - 1, b.cnt) = b.seqnum - 1;