[hibernate-dev] Memory consumption

[Guillaume Smet]

On Thu, May 17, 2012 at 6:53 PM, Andrej Golovnin <golovnin at gmx.net> wrote:
> Query: SELECT * FROM test WHERE oid IN (100, 101, 102, 100, 100, 100);
> "Index Scan using test_pkey on public.test  (cost=0.00..29.72 rows=*6* width=9) (actual time=0.024..0.034 rows=3 loops=1)"
>
> Query: SELECT * FROM test WHERE oid IN (100, 101, 102, null, null, null);
> "Index Scan using test_pkey on public.test  (cost=0.00..29.69 rows=*3* width=9) (actual time=0.023..0.033 rows=3 loops=1)"
>
> As you see the difference is only in the calculated costs.

And in the estimation of the number of rows returned. At least on
PostgreSQL, the last query is more accurate because the estimated row
count is what we want.

When you have a lot of joins, even a slight difference in an estimated
row count can lead to quite different plans.

> If you want, I could show on monday explain output from MySQL 5.5 and Oracle 11g.

Might be interesting.