Horton

Re: Optimization issue of branching UNION ALL

Andrei Lepikhov <a.lepikhov@postgrespro.ru>

View thread View on postgr.es
From: Andrey Lepikhov <a.lepikhov@postgrespro.ru>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2022-12-22T11:59:41Z
Lists: pgsql-hackers 
On 22/12/2022 06:50, Tom Lane wrote:
>> 2. Iterative passes along the append_rel_list for replacing vars in the
>> translated_vars field. I can't grasp real necessity of passing all the
>> append_rel_list during flattening of an union all leaf subquery. No one
>> can reference this leaf, isn't it?
> 
> After thinking about that for awhile, I believe we can go further:
> the containing_appendrel is actually the *only* part of the upper
> query that needs to be adjusted.  So we could do something like
> the attached.
> 
> This passes check-world, but I don't have quite enough confidence
> in it to just commit it.
Thanks, I have written the letter because of some doubts too. But only 
one weak point I could imagine - if someday sql standard will be changed.
Your code looks better, than previous attempt.

-- 
regards,
Andrey Lepikhov
Postgres Professional

Commits

  1. Avoid O(N^2) cost when pulling up lots of UNION ALL subqueries.

  2. Add some recursion and looping defenses in prepjointree.c.