Re: Optimizer isn't perfect - Mailing list pgsql-general

From Tom Lane
Subject Re: Optimizer isn't perfect
Date
Msg-id 27639.1062173464@sss.pgh.pa.us
Whole thread Raw
In response to Optimizer isn't perfect  (Greg Stark <gsstark@mit.edu>)
List pgsql-general
Greg Stark <gsstark@mit.edu> writes:
> This is the same thing someone else pointed out a while ago. The consequences
> of guessing wrong and favouring a full table scan instead of index lookups
> with nested loops are simply much more severe than the reverse.

It is easy to demonstrate cases where this isn't true.

            regards, tom lane

pgsql-general by date:

Previous
From: "Jenny -"
Date:
Subject: LOCK.tag(figuring out granularity of lock)--
Next
From: Sander Smeenk
Date:
Subject: Re: PostgreSQL upgrade -> fails to start server