Re: PostgreSQL upgraded to 8.2 but forcing index scan on query produces faster

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Christian Brink <cbrink(at)r-stream(dot)com>
Cc: pgsql-performance(at)postgresql(dot)org
Subject: Re: PostgreSQL upgraded to 8.2 but forcing index scan on query produces faster
Date: 2010-03-22 19:21:35
Message-ID: 12444.1269285695@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Christian Brink <cbrink(at)r-stream(dot)com> writes:
> -> Nested Loop (cost=0.01..2416.59 rows=22477 width=4)
> (actual time=0.595..2.150 rows=225 loops=1)
> -> Index Scan using sysstrings_pkey on sysstrings
> (cost=0.00..8.27 rows=1 width=182) (actual time=0.110..0.112 rows=1 loops=1)
> Index Cond: (id = 'net/Console/Employee/Day End
> Time'::text)
> -> Index Scan using sales_tranzdate_index on sales s
> (cost=0.01..1846.40 rows=22477 width=12) (actual time=0.454..1.687
> rows=225 loops=1)
> Index Cond: ((s.tranzdate >= ('2010-02-15'::date +
> (sysstrings.data)::time without time zone)) AND (s.tranzdate <
> ('2010-02-16'::date + (sysstrings.data)::time without time zone)))
> Filter: ((NOT void) AND (NOT suspended))

The fundamental reason why you're getting a bad plan choice is the
factor-of-100 estimation error here. I'm not sure you can do a whole
lot about that without rethinking the query --- in particular I would
suggest trying to get rid of the non-constant range bounds. You're
apparently already plugging in an external variable for the date,
so maybe you could handle the time of day similarly instead of joining
to sysstrings for it.

regards, tom lane

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Pierre C 2010-03-22 20:55:27 Re: Block at a time ...
Previous Message Christian Brink 2010-03-22 19:09:11 PostgreSQL upgraded to 8.2 but forcing index scan on query produces faster