From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | "Kevin Grittner" <Kevin(dot)Grittner(at)wicourts(dot)gov> |
Cc: | pgsql-hackers(at)postgreSQL(dot)org |
Subject: | Re: Fixing Grittner's planner issues |
Date: | 2009-02-05 17:37:13 |
Message-ID: | 21640.1233855433@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
"Kevin Grittner" <Kevin(dot)Grittner(at)wicourts(dot)gov> writes:
> Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> If it referred to any variables of the left side of the upper join
>> (ie, "C" here) then we couldn't convert it to a separate join at
> all.
> The only thing that comes to mind for me that seems possibly helpful
> is that we have typically considered it obvious that in the context of
> the NOT EXISTS clause we have already established that ("CD"."caseNo"
> = "C"."caseNo") AND ("CD"."countyNo" = "C"."countyNo") and have not
> been at all consistent about whether we used C or CD to compare to
> CD2. Our operating assumption has been that it didn't matter in that
> context.
Yeah, I had thought about suggesting that you could use that to
determine which type of plan got chosen, but immediately dismissed it
as too bletcherous for words. In any case the same type of problem
could occur in scenarios where the upper and lower join conditions
aren't so neatly related.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Heikki Linnakangas | 2009-02-05 19:54:50 | Re: Hot standby, recovery infra |
Previous Message | Kevin Grittner | 2009-02-05 17:29:52 | Re: Fixing Grittner's planner issues |