Re: Rowcounts marked by create_foreignscan_path() - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Rowcounts marked by create_foreignscan_path()
Date
Msg-id 12798.1392692600@sss.pgh.pa.us
Whole thread Raw
In response to Rowcounts marked by create_foreignscan_path()  (Etsuro Fujita <fujita.etsuro@lab.ntt.co.jp>)
Responses Re: Rowcounts marked by create_foreignscan_path()  (Etsuro Fujita <fujita.etsuro@lab.ntt.co.jp>)
List pgsql-hackers
Etsuro Fujita <fujita.etsuro@lab.ntt.co.jp> writes:
> Why does create_foreignscan_path() not set the rowcounts based on
> ParamPathInfo when the path is a parameterized path?

The calling FDW is supposed to do that; note the header comment.
I'm not sure that it'd be an improvement to change the API spec
to be "create_foreignscan_path has no intelligence, except that
sometimes it will decide to override your rows estimate anyway;
nonetheless, it takes your cost estimate as gospel".
        regards, tom lane



pgsql-hackers by date:

Previous
From: Etsuro Fujita
Date:
Subject: Rowcounts marked by create_foreignscan_path()
Next
From: Etsuro Fujita
Date:
Subject: Re: Rowcounts marked by create_foreignscan_path()