On Sun, Mar 12, 2017 at 9:50 AM, SuperCiccio wrote:
> But even if I create a specific index
> CREATE INDEX on datatable (((jsonfield#>>'{path1,path2}')::numeric));
> it isn't used in such a query; it is used in this query:
> select field1,field2 from datatable where
> (jsonfield#>>'{path1,path2}')
This is a kind of document, there are many other kinds, with different
json structure. So, a general index would be better.
But even if I create a specific index
CREATE INDEX on datatable (((jsonfield#>>'{path1,path2}')::numeric));
it isn't used in such a query; it is used in this query:
selec
On 3/11/2017 3:55 PM, SuperCiccio wrote:
Is there a way to make the following query use some sort of indexing?
select field1,field2 from datatable where
(jsonfield#>>'{path1,path2}')::numeric < 1000;
if the '{path1,path2}' part is constant, I'd think you could create an
index on (jsonfield #
Is there a way to make the following query use some sort of indexing?
select field1,field2 from datatable where
(jsonfield#>>'{path1,path2}')::numeric < 1000;
Thanks
On Mon, 19 Jul 2004, Michal Hlavac wrote:
> hello,
>
> I have in db table with 3 columns... table name is l_model_to_part
> columns: i_model_id, i_part_id, i_year
>
> I have index on every column separately and primary key is (i_model_id,
> i_part_id, i_year)
>
> when I use: EXPLAIN SELECT * FROM
Michal Hlavac wrote:
> my question is, why postgres doesn't use index_50 in second query???
> when I create this structure in clear db, everything is all right...
>
of course, index_50 is BTREE index with one column (i_model_id)...
thanx, hlk
--
[ miso hlavac ][ [EMAIL PROTECTED] ][ http://hlavki.s
hello,
I have in db table with 3 columns... table name is l_model_to_part
columns: i_model_id, i_part_id, i_year
I have index on every column separately and primary key is (i_model_id,
i_part_id, i_year)
when I use: EXPLAIN SELECT * FROM l_model_to_part WHERE i_part_id=234;
