strange slow query performance - Mailing list pgsql-general

From Ben Snaidero
Subject strange slow query performance
Date
Msg-id CAEPE5bOdhQm41PEjvTeQK2mrUA5xeOnZ-Oe1Kw=AShvfEbAZpw@mail.gmail.com
Whole thread Raw
Responses Re: strange slow query performance
Re: strange slow query performance
List pgsql-general
The following query runs as expected.

explain analyze SELECT MainTable.gid AS MainTable_gid,MainTable.datetime AS MainTable_datetime,MainTable.objectid AS MainTable_objectid,MainTable.type AS MainTable_type FROM MainTable  
   WHERE objectid = ANY(ARRAY(SELECT NE.objectid AS entityId FROM ( nodeobject NE INNER JOIN (       Node N1NE INNER JOIN Node N2NE   ON ((N1NE.iLeft >= N2NE.iLeft) AND (N1NE.iLeft <= N2NE.iRight)))   ON NE.inodeid = N1NE.iId ) WHERE N2NE.iId = ANY(ARRAY[14438,10814]))) 
     AND objectid=3161;
                                                                               QUERY PLAN                                                                                
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 Index Scan using ix_MainTable_objectid_datetime on MainTable  (cost=3254.91..3264.39 rows=1 width=32) (actual time=33.094..33.094 rows=0 loops=1)
   Index Cond: ((objectid = ANY ($3)) AND (objectid = 3161))
   InitPlan 1 (returns $3)
     ->  Nested Loop  (cost=1.10..3254.75 rows=16296 width=4) (actual time=0.403..26.147 rows=19042 loops=1)
           ->  Nested Loop  (cost=0.81..1784.15 rows=618 width=4) (actual time=0.249..10.485 rows=2082 loops=1)
                 ->  Index Scan using pk_node on node n2ne  (cost=0.41..16.80 rows=2 width=8) (actual time=0.103..0.233 rows=2 loops=1)
                       Index Cond: (iid = ANY ('{14438,10814}'::integer[]))
                 ->  Index Only Scan using ix_node on node n1ne  (cost=0.41..880.59 rows=309 width=8) (actual time=0.074..5.022 rows=1041 loops=2)
                       Index Cond: ((ileft >= n2ne.ileft) AND (ileft <= n2ne.iright))
                       Heap Fetches: 30
           ->  Index Only Scan using ix_nodeobject_inodeid_objectid on nodeobject ne  (cost=0.29..1.81 rows=57 width=8) (actual time=0.002..0.007 rows=9 loops=2082)
                 Index Cond: (inodeid = n1ne.iid)
                 Heap Fetches: 13973
 Planning time: 5.693 ms
 Execution time: 33.383 ms
(15 rows)

But when we add a second condition to the where clause it causes the query performance to drop dramatically.  Values in "any(array" do not make a difference.  In this example they are the same but even for different values performance is still the poor.  Any ideas as to why this is happening?

explain analyze SELECT MainTable.gid AS MainTable_gid,MainTable.datetime AS MainTable_datetime,MainTable.objectid AS MainTable_objectid,MainTable.type AS MainTable_type FROM MainTable  
   WHERE objectid = ANY(ARRAY(SELECT NE.objectid AS entityId FROM ( nodeobject NE INNER JOIN (       Node N1NE INNER JOIN Node N2NE   ON ((N1NE.iLeft >= N2NE.iLeft) AND (N1NE.iLeft <= N2NE.iRight)))   ON NE.inodeid = N1NE.iId ) WHERE N2NE.iId = ANY(ARRAY[14438,10814]))) 
     AND objectid = ANY(ARRAY(SELECT NE.objectid AS entityId FROM ( nodeobject NE INNER JOIN (       Node N1NE INNER JOIN Node N2NE   ON ((N1NE.iLeft >= N2NE.iLeft) AND (N1NE.iLeft <= N2NE.iRight)))   ON NE.inodeid = N1NE.iId ) WHERE N2NE.iId = ANY(ARRAY[14438,10814]))) AND objectid=3161;

                                                                                QUERY PLAN                                                                                
 
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-
 Index Scan using ix_MainTable_objectid_datetime on MainTable  (cost=6509.66..6534.02 rows=1 width=32) (actual time=16442.004..16442.004 rows=0 loops=1)
   Index Cond: ((objectid = ANY ($3)) AND (objectid = ANY ($7)) AND (objectid = 3161))
   InitPlan 1 (returns $3)
     ->  Nested Loop  (cost=1.10..3254.75 rows=16296 width=4) (actual time=0.438..28.484 rows=19042 loops=1)
           ->  Nested Loop  (cost=0.81..1784.15 rows=618 width=4) (actual time=0.172..7.141 rows=2082 loops=1)
                 ->  Index Scan using pk_node on node n2ne  (cost=0.41..16.80 rows=2 width=8) (actual time=0.046..0.111 rows=2 loops=1)
                       Index Cond: (iid = ANY ('{14438,10814}'::integer[]))
                 ->  Index Only Scan using ix_node on node n1ne  (cost=0.41..880.59 rows=309 width=8) (actual time=0.073..3.438 rows=1041 loops=2)
                       Index Cond: ((ileft >= n2ne.ileft) AND (ileft <= n2ne.iright))
                       Heap Fetches: 30
           ->  Index Only Scan using ix_nodeobject_inodeid_objectid on nodeobject ne  (cost=0.29..1.81 rows=57 width=8) (actual time=0.003..0.010 rows=9 loops=2082)
                 Index Cond: (inodeid = n1ne.iid)
                 Heap Fetches: 13973
   InitPlan 2 (returns $7)
     ->  Nested Loop  (cost=1.10..3254.75 rows=16296 width=4) (actual time=0.056..11.786 rows=19042 loops=1)
           ->  Nested Loop  (cost=0.81..1784.15 rows=618 width=4) (actual time=0.034..1.343 rows=2082 loops=1)
                 ->  Index Scan using pk_node on node n2ne_1  (cost=0.41..16.80 rows=2 width=8) (actual time=0.007..0.016 rows=2 loops=1)
                       Index Cond: (iid = ANY ('{14438,10814}'::integer[]))
                 ->  Index Only Scan using ix_node on node n1ne_1  (cost=0.41..880.59 rows=309 width=8) (actual time=0.012..0.581 rows=1041 loops=2)
                       Index Cond: ((ileft >= n2ne_1.ileft) AND (ileft <= n2ne_1.iright))
                       Heap Fetches: 30
           ->  Index Only Scan using ix_nodeobject_inodeid_objectid on nodeobject ne_1  (cost=0.29..1.81 rows=57 width=8) (actual time=0.002..0.004 rows=9 loops=2082)
                 Index Cond: (inodeid = n1ne_1.iid)
                 Heap Fetches: 13973
 Planning time: 4.860 ms
 Execution time: 16442.462 ms
(26 rows)


Thanks in advance
Ben.



pgsql-general by date:

Previous
From: Nicola Contu
Date:
Subject: Re: pgbouncer
Next
From: "David G. Johnston"
Date:
Subject: Re: pgbouncer