I have a new summary table built againt a single data table. Simple
select, count(*) and group by type. After I built it, I found that the
optimizer will choose to use it for existing queries IF there is no
order by in query. Has anyone read that this is a limitation?
create summary table s_t as (select f1, f2, count(*) from t1 group by
f1, f2)
select distinct f1, f2 from t1
- explain shows optimizer uses s_t
select distinct f1, f2 from t1 order by f1, f2
- does not use s_t
Comments?
