Re: order by slowing down a query by 80 times

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



"Kevin Grittner" <Kevin.Grittner@xxxxxxxxxxxx> writes:
> Rajesh Kumar Mallah <mallah.rajesh@xxxxxxxxx> wrote:
>> just by removing the order by co_name reduces the query time
>> dramatically from  ~ 9 sec  to 63 ms. Can anyone please help.
 
> The reason is that one query allows it to return *any* 25 rows,
> while the other query requires it to find a *specific* set of 25
> rows.  It happens to be faster to just grab any old set of rows than
> to find particular ones.

I'm guessing that most of the cost is in repeated evaluations of the
filter clause
	(co_name_vec)::tsvector @@ to_tsquery('manufacturer'::text)

There are two extremely expensive functions involved there (cast to
tsvector and to_tsquery) and they're being done over again at what
I suspect is practically every table row.  The unordered query is
so much faster because it stops after having evaluated the text
search clause just a few times.

The way to make this go faster is to set up the actually recommended
infrastructure for full text search, namely create an index on
(co_name_vec)::tsvector (either directly or using an auxiliary tsvector
column).  If you don't want to maintain such an index, fine, but don't
expect full text search queries to be quick.

			regards, tom lane

-- 
Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance


[Postgresql General]     [Postgresql PHP]     [PHP Users]     [PHP Home]     [PHP on Windows]     [Kernel Newbies]     [PHP Classes]     [PHP Books]     [PHP Databases]     [Yosemite]

  Powered by Linux