Le lundi 08 octobre 2007, Dmitry Koterov a écrit :
> explain analyze
> SELECT id FROM "table" WHERE name LIKE 'dt%';
>
> having a btree index on "name" column. But unfortunately it uses seqscan
> instead of index scan, it's too slow.
It seems to me you'd benefit from reading this page of the fine manual:
http://www.postgresql.org/docs/current/interactive/indexes-opclass.html
Regards,
--
dim