Is it a known bug if optimizer, that Postgres can't
use indexes created on a int8 (bigint) ?
This problem exists in 7.1.3 and even in
7.2.
test=# CREATE TABLE parent_child
(
test(# parent_id int8 NOT NULL, test(# parent_type int4 NOT NULL, test(# child_id int8 NOT NULL PRIMARY KEY, test(# child_type int4 NOT NULL, test(# account_id int8 NOT NULL, test(# p_begin date, test(# suspended int4); NOTICE: CREATE TABLE / PRIMARY KEY will create implicit index 'parent_child_pkey' for table 'parent_child' CREATE test=# explain select * from parent_child where child_id=1; NOTICE: QUERY PLAN: Seq Scan on parent_child (cost=0.00..22.50
rows=1 width=40)
EXPLAIN
test=# create index parent_child_childtype on parent_child(child_type); CREATE test=# explain select * from parent_child where child_type=1; NOTICE: QUERY PLAN: Index Scan using parent_child_childtype on
parent_child (cost=0.00..17.07 rows=5 width=40)
EXPLAIN
test=# |
- Re: [BUGS] BUG: bigint and indexes ? Victor Krasinsky
- Re: [BUGS] BUG: bigint and indexes ? Stephan Szabo