Hello all,
While testing Index scan,I found the following phonomenon.
SELECT id from xxxxxx
where id=10 or id=11;
is very fast.
But
SELECT id from xxxxxx
where (id>=10 and id<=10)
or (id>=11 and id<=11);
is very slow.
Why ?
The EXPLAIN(not verbose) output of both SQL are same
except cost and size.
NOTICE: QUERY PLAN:
Index Scan using xxxxxx_pkey, xxxxxx_pkey on xxxxxx
(cost=1136.17 size=197 width=4)
It seems that (id>=..) is included in indexqual but (id<=.. )
is not.
Thanks.
Hiroshi Inoue
Inoue(at)tpf(dot)co(dot)jp