Sebastian Böck wrote:
Hello,

if i have the following (simple) table layout:

create table a (
  id serial primary key
);

create table b (
  id integer references a,
  test text
);

create view c as
  select a.id,b.test from a
  left join b
  on a.id = b.id;

test=# EXPLAIN SELECT * from g;

test=# EXPLAIN SELECT id from g;

so the only difference is the width estimation.

But why is the scan on table b performed?
If i understand it correctly this is unnecessary because the
result contains only rows from table a.

It's only unnecessary in the case where there is a 1:1 correspondence between a.id and b.id - if you had more than one matching row in "b" then there'd be repeated rows from "a" in the result. Not sure if PG can tell what the situation is regarding references and pkeys, but in your example you don't have one anyway.


--
  Richard Huxton
  Archonet Ltd

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org

Reply via email to