Horton

Re: make add_paths_to_append_rel aware of startup cost

Andy Fan <zhihuifan1213@163.com>

View thread View on postgr.es
From: Andy Fan <zhihuifan1213@163.com>
To: David Rowley <dgrowleyml@gmail.com>
Cc: Thomas Munro <thomas.munro@gmail.com>, Andy Fan <zhihui.fan1213@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>, Zhang Mingli <zmlpostgres@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2024-02-16T01:19:05Z
Lists: pgsql-hackers 
David Rowley <dgrowleyml@gmail.com> writes:

> I'd be more happy using this one as percentage-wise, the cost
> difference is much larger.

+1 for the percentage-wise.
>
> I checked that the t2.thounsand = 0 query still tests the cheap
> startup paths in add_paths_to_append_rel().

I get the same conclusion here. 
>
> So, if nobody has any better ideas, I'm just going to push the " and
> t2.thousand = 0" adjustment.

LGTM.

-- 
Best Regards
Andy Fan

Commits

  1. Attempt to stabilize flapping regression test

  2. Consider cheap startup paths in add_paths_to_append_rel