| From: | David Rowley <dgrowleyml(at)gmail(dot)com> |
|---|---|
| To: | Денис Смирнов <darthunix(at)gmail(dot)com> |
| Cc: | PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Use virtual tuple slot for Unique node |
| Date: | 2023-08-30 21:59:59 |
| Message-ID: | CAApHDvo1D1++JsxPbzA5kpsesZD8vv=DPk=OBAjVUpoAYyaW0w@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Thu, 31 Aug 2023 at 05:37, Денис Смирнов <darthunix(at)gmail(dot)com> wrote:
> I have inspected the performance of the GROUP BY and DISTINCT queries for the sorted data streams and found out, that Group node (produced by GROUP BY) works faster then the Unique node (produced by DISTINCT). The flame graph should out the reason - Unique palloc`s tuples for the result slot while the Group node doesn’t.
>
> I wonder, why do we use minimal tuples for the Unique node instead of the virtual ones? It looks like there is no actual reason for that as Unique doesn’t make any materialization.
It would be good to see example queries and a demonstration of the
performance increase. I'm not disputing your claims, but showing some
performance numbers might catch the eye of a reviewer more quickly.
You should also add this to the September commitfest at
https://commitfest.postgresql.org/44/
David
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2023-08-30 22:19:26 | Re: Query execution in Perl TAP tests needs work |
| Previous Message | Nathan Bossart | 2023-08-30 21:56:48 | Re: pg_stat_get_backend_subxact() and backend IDs? |