On Tue, 9 Apr 2002, Alexis Maldonado wrote:
> Ok.. I know its provably something im doing dumb..
> but here it goes..
>
> I have 2 tables that are the same:
>
> "temp_table" and "table"
>
> "temp _table" has 7,761 rows and "table" is empty
>
> the columns for both tables are: ID (primary key sequence), index, column1,
> column2
>
> when i run:
>
> Insert Into table
> select index, column1, column2
> from temp_table
> where index NOT IN (select index from table)
IN is unfortunately implemented slowly (I think the FAQ answer has more
details)
You can often get better performance using exists, I think the equivalent
would be:
insert into table
select index, column1, column2 from temp_table
where NOT EXISTS (select * from table where table.index=temp_Table.index)
