Re: Trying to eliminate union and sort

From: Brian Fehrle <brianf(at)consistentstate(dot)com>
To: pgsql-performance(at)postgresql(dot)org
Subject: Re: Trying to eliminate union and sort
Date: 2013-07-15 16:12:20
Message-ID: 51E41F64.7090102@consistentstate.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On 07/12/2013 04:43 PM, Josh Berkus wrote:
>> As for the counts on the tables:
>> table1 3,653,472
>> table2 2,191,314
>> table3 25,676,589
>>
>> I think it's safe to assume right now that any resulting joins are not
>> one-to-one
> Hmmm? How is doing a subselect in the SELECT clause even working, then?
>
Oh my, this is sad. the query in all returns 9,955,729 rows, so the sub
queries are run on each of these resulting rows, however in this entire
result set, subquery 1 returns 16 distinct rows, subquery 2 returns 63
different rows, but those sub queries are run over 10 million times to
return these few distinct rows. So it's running many times, but
returning the same small set of data over and over again.

- Brian F

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Marcos Luis 2013-07-15 17:00:06 Thought you'd find this interesting
Previous Message Sergey Konoplev 2013-07-15 04:33:57 Re: how to speed up the index creation in GP?