Re: planner chooses incremental but not the best one - Mailing list pgsql-hackers

From Tom Lane
Subject Re: planner chooses incremental but not the best one
Date
Msg-id 3783538.1702852425@sss.pgh.pa.us
Whole thread Raw
In response to Re: planner chooses incremental but not the best one  (Tomas Vondra <tomas.vondra@enterprisedb.com>)
List pgsql-hackers
Tomas Vondra <tomas.vondra@enterprisedb.com> writes:
> Yeah, seems like that's the right thing to do. FWIW I've been often
> confused by these fields, because we use tuples and rows as synonyms,
> but in this particular case that's not the same. I wonder if this is
> just a manifestation of this confusion.

For tables, one is the raw number of rows on-disk and the other is the
number of rows predicted to pass the relation's quals.  For something
like an appendrel that doesn't enforce any quals (today anyway), they
should probably be the same; but you need to be sure you're adding
up the right numbers from the inputs.

            regards, tom lane



pgsql-hackers by date:

Previous
From: Tomas Vondra
Date:
Subject: Re: planner chooses incremental but not the best one
Next
From: Andy Fan
Date:
Subject: Re: Is a clearer memory lifespan for outerTuple and innerTuple useful?