Re: BUG #9896: Bug in FULL OUTER JOIN

From: Vik Fearing <vik(dot)fearing(at)dalibo(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-bugs(at)postgresql(dot)org
Subject: Re: BUG #9896: Bug in FULL OUTER JOIN
Date: 2014-04-07 20:39:28
Message-ID: 53430D00.9040009@dalibo.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

On 04/07/2014 10:17 PM, Tom Lane wrote:
> I wrote:
>> sqlpro(at)sqlspot(dot)com writes:
>>> SELECT *
>>> FROM T_CLIENT_CLI AS C
>>> FULL OUTER JOIN T_PROSPECT_PSP AS P
>>> ON C.CLI_SIREN = P.PSP_SIREN
>>> OR C.CLI_ENSEIGNE = P.PSP_ENSEIGNE;
>> Can you show us the query plans used by those systems?
> For the archives' sake: the OP sent me a not-too-useful screen shot
> in which SQL Server claims it's using a merge join for this query.
>
> I find this less than credible: what linear sort order would bring
> together all the potentially joinable rows? If they're getting the
> right answer at all, there must be some secret sauce in there someplace.
> Perhaps they're just Doing It The Hard Way with state storage
> proportional to the size of the relations?

And Oracle does it in two steps: http://www.sqlfiddle.com/#!4/607e0/1/0

--
Vik

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message fburgess 2014-04-07 22:25:14 Re: Configuring Standby Server in PostgreSQL 9.3.3
Previous Message Tom Lane 2014-04-07 20:17:58 Re: BUG #9896: Bug in FULL OUTER JOIN