Wrong actual number of rows in the Query Plan

From: Vahe Evoyan <vahe(dot)evoyan(at)gmail(dot)com>
To: pgsql-performance(at)postgresql(dot)org
Subject: Wrong actual number of rows in the Query Plan
Date: 2013-02-21 07:24:27
Message-ID: CAETCRjg_3kp-MniE6ZmEH4ZMSE4xiSD=01vsuXrQae4Yxwq32Q@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Hello,

The Query Plan for the query below shows a large number in its actual rows
count by an unknown reason. As a result Merge Join works on a large enough
data to slow down the query.

The table which I query has the following description:

Table
"public.qor_value"
Column | Type |
Modifiers | Storage | Description
-------------+------------------------+---------------------------------------------------------------------+----------+-------------
value_id | integer | not null default
nextval('qor_value_denorm_value_id_seq'::regclass) | plain |
run_id | integer | not
null | plain
|
dft_id | integer | not
null | plain
|
stat_id | integer | not
null | plain
|
key | character varying(128)
| |
extended |
value | numeric(22,10)
| |
main |
line_number | integer | not null default
nextval('qor_value_line_numbering'::regclass) | plain |
file_number | integer | not
null | plain
|
Indexes:
"qor_value_cluster" btree (run_id, stat_id) CLUSTER INVALID
"qor_value_filtered_self_join" btree (run_id, stat_id, key, dft_id,
line_number) INVALID
"qor_value_self_join" btree (run_id, stat_id, dft_id, key, line_number)

Here is the query:

EXPLAIN ANALYZE
SELECT *
FROM "qor_value" V1
INNER JOIN "qor_value" V2
USING ("dft_id", "stat_id", "key")
WHERE
V1."stat_id" = 342 AND
V1."run_id" = 60807 AND
V2."run_id" = 60875;

QUERY
PLAN
-----------------------------------------------------------------------------------------------------------------------------------------------------
Merge Join (cost=0.00..2513.96 rows=1 width=72) (actual
time=127.361..473.687 rows=66460 loops=1)
Merge Cond: ((v1.dft_id = v2.dft_id) AND ((v1.key)::text =
(v2.key)::text))
-> Index Scan using qor_value_self_join on qor_value v1
(cost=0.00..1255.60 rows=275 width=51) (actual time=89.549..97.045
rows=1388 loops=1)
Index Cond: ((run_id = 60807) AND (stat_id = 342))
-> Index Scan using qor_value_self_join on qor_value v2
(cost=0.00..1255.60 rows=275 width=51) (actual time=37.796..134.286
rows=66343 loops=1)
Index Cond: ((run_id = 60875) AND (stat_id = 342))
Total runtime: 544.646 ms
(7 rows)

Note that the second Index Scan has 66343 rows in place of 1388. Here is
the query which proves that:

SELECT COUNT(*) FROM "qor_value" WHERE run_id = 60875 AND stat_id = 342;
count
-------
1388

Please help me to figure out where the problem is.
Thanks in advance,
Vahe

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Mark Smith 2013-02-21 09:59:01 Poor performance after update from SLES11 SP1 to SP2
Previous Message Josh Berkus 2013-02-21 03:14:10 Re: High CPU usage / load average after upgrading to Ubuntu 12.04