Horton

Re: Introduce new multi insert Table AM and improve performance of various SQL commands with it for Heap AM

Jingtang Zhang <mrdrivingduck@gmail.com>

View thread
From: Jingtang Zhang <mrdrivingduck@gmail.com>
To: Daniil Davydov <3danissimo@gmail.com>
Cc: pgsql-hackers@lists.postgresql.org, PostgreSQL-development <pgsql-hackers@postgresql.org>, Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Date: 2025-04-06T13:54:44Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. libpq: Fix some issues in TAP tests for service files

  2. Multiple revisions to the GROUP BY reordering tests

  3. Explore alternative orderings of group-by pathkeys during optimization.

Hi, Daniil Davydov.

Thanks for sharing your patch!

It was quite a while since I last looked at the patch. I've tested it again,
and still get regression on patched version where a table has many columns.
And it is totally CPU-bounded on tts_virtual_copyslot.

Unpatched version:
1 col:
Time: 8909.714 ms (00:08.910)
Time: 8803.579 ms (00:08.804)
Time: 8600.415 ms (00:08.600)
32 cols:
Time: 12911.699 ms (00:12.912)
Time: 13543.491 ms (00:13.543)
Time: 13325.368 ms (00:13.325)

Patched version:
1 col:
Time: 3532.841 ms (00:03.533)
Time: 3598.223 ms (00:03.598)
Time: 3515.858 ms (00:03.516)
32 cols:
Time: 35647.724 ms (00:35.648)
Time: 35596.233 ms (00:35.596)
Time: 35669.106 ms (00:35.669)

I've tested your patch with tuplestore and found the regression does not exist
anymore, but I haven't look deep inside it.

Patched version (with tuplestore):
1 col:
Time: 3500.502 ms (00:03.501)
Time: 3486.886 ms (00:03.487)
Time: 3514.233 ms (00:03.514)
32 cols:
Time: 10375.391 ms (00:10.375)
Time: 10248.256 ms (00:10.248)
Time: 10248.289 ms (00:10.248)

It seems to be a good idea if there is no other issue with your patch.

--
Regards,
Jingtang