On 23 avr, 14:12, st...@xxxxxxxxxxxxxxxx (Gregory Stark) wrote: > <mate...@xxxxxxxxxxxxxx> writes: > > This is the result of an EXPLAIN: > >... > > I suppose that the problem comes from the Bitmap Heap Scan which costs > > a lot, but I can't be totally sure. > > > Any idea on where I should be investigating ? > > Try posting an EXPLAIN ANALYZE which will actually run the query and include > information to compare against the estimates. > > -- > Gregory Stark > EnterpriseDB http://www.enterprisedb.com > Ask me about EnterpriseDB's PostGIS support! > > -- > Sent via pgsql-general mailing list (pgsql-gene...@xxxxxxxxxxxxxx) > To make changes to your subscription:http://www.postgresql.org/mailpref/pgsql-general Thanks for your answer. Here is the EXPLAIN ANALYZE of this query: Limit (cost=392.43..392.43 rows=1 width=4) (actual time=689.016..689.017 rows=1 loops=1)" -> Sort (cost=392.22..392.43 rows=83 width=4) (actual time=688.552..688.811 rows=1000 loops=1)" Sort Key: flg_rid" -> Index Scan using prj_frm_flg_pkey on prj_frm_flg (cost=0.00..389.58 rows=83 width=4) (actual time=0.186..687.578 rows=1000 loops=1)" Index Cond: (flg_mid = 3)" Filter: ((NOT flg_fav) AND (NOT flg_notif) AND (NOT flg_post))" Total runtime: 689.092 ms"
