On 02/18/2018 06:37 AM, David Rowley wrote:
> On 18 February 2018 at 12:35, hmidi slim <hmidi.slim2@xxxxxxxxx> wrote:
>> Is there an other optimized solution to make a query such this:
>> select * from (
>> select e.name, e1.name, e.id
>> from establishment as e, establishment as e1
>> where e.id <> e1.id
>> and e1.id = 1
>> and ST_DWithin(geom, ST_MakePoint(e1.longitude, e1.latitude)::geography,
>> 1000)) as tmp inner join product as p on p.establishment_id = tmp.id
>
> You really should state what is wrong with the performance of the
> above version and also include the EXPLAIN (ANALYZE, BUFFERS) of that
> query.
>
> The version of PostgreSQL that you're running it on is also a good
> thing to share.
>
> Details of the indexes which you've defined on the tables are also
> useful information. It would especially be good to know if you've
> added an index on product (establishment_id), for example.
>
There's actually a wiki page with overview of what should be included in
"slow query" questions:
https://wiki.postgresql.org/wiki/Slow_Query_Questions
regards
--
Tomas Vondra http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
