Re: Problem using LIMIT 1 when only 1 record exists. - Mailing list pgsql-bugs

From Tom Lane
Subject Re: Problem using LIMIT 1 when only 1 record exists.
Date
Msg-id 1730.1070058198@sss.pgh.pa.us
Whole thread Raw
In response to Re: Problem using LIMIT 1 when only 1 record exists.  (Neil Conway <neilc@samurai.com>)
List pgsql-bugs
Neil Conway <neilc@samurai.com> writes:
> "Steve Thames" <steve@softlife.com> writes:
>> The SQL command:
>> SELECT last FROM table WHERE symbol='Symbol' AND expmoyr='Mmm-YY' ORDER BY
>> qdate DESC LIMIT 1;
>>
>> This query works fine when there is more than one record meeting the
>> criteria.  When there is only 1, the query locks and no result is
>> returned.

> I'm skeptical that LIMIT is broken. Can you provide a reproducible
> example?

I'll bet very long odds that "locks" should be read as "ran longer than
Steve wanted to wait".  Before this 7.4 change:

2002-11-22 17:10  tgl

    * src/: backend/executor/nodeLimit.c, include/nodes/execnodes.h:
    Redesign internal logic of nodeLimit so that it does not need to
    fetch one more row from the subplan than the COUNT would appear to
    require.  This costs a little more logic but a number of people
    have complained about the old implementation.

nodeLimit would have tried to fetch two rows from the underlying plan.
If only one matching row is actually available, this would result in
scanning the whole source table.

            regards, tom lane

pgsql-bugs by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: [ADMIN] Misplaced modifier in Postgresql license
Next
From: Bruce Momjian
Date:
Subject: Re: [GENERAL] [ADMIN] Misplaced modifier in Postgresql license