INM1=# explain select ref_num from grls where ag_id in (24);
Index Scan using grls_ag_id on grls (cost=0.00..597.87 rows=849 width=12)
IMN1=# explain select ref_num from grls where ag_id in (select 24);
Seq Scan on grls (cost=0.00..992.44 rows=30195 width=12)
SubPlan
-> Materialize (cost=0.00..0.00 rows=0 width=0)
-> Result (cost=0.00..0.00 rows=0 width=0)
IMN1=# explain select ref_num from grls where ag_id = (select 24);
Index Scan using grls_ag_id on grls (cost=0.00..597.87 rows=849 width=12)
InitPlan
-> Result (cost=0.00..0.00 rows=0 width=0)
". . . in (24) " work in 35 faster than ". . . in (select 24)". In the
last don't use indices for increase perfomans.
It's been normaly work in PostgreSQL version 7.2 ? Or not?
P.S.
Sorry my English.