Horton

Re: pg_stat_statements and "IN" conditions

Dmitry Dolgov <9erthalion6@gmail.com>

View thread
From: Dmitry Dolgov <9erthalion6@gmail.com>
To: Yasuo Honda <yasuo.honda@gmail.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Peter Smith <smithpb2250@gmail.com>, vignesh C <vignesh21@gmail.com>, Michael Paquier <michael@paquier.xyz>, Nathan Bossart <nathandbossart@gmail.com>, "Gregory Stark (as CFM)" <stark.cfm@gmail.com>, David Geier <geidav.pg@gmail.com>, Sergei Kornilov <sk@zsrv.org>, Alvaro Herrera <alvherre@alvh.no-ip.org>, Marcos Pegoraro <marcos@f10.com.br>, Robert Haas <robertmhaas@gmail.com>, Zhihong Yu <zyu@yugabyte.com>, David Steele <david@pgmasters.net>, PostgreSQL-development <pgsql-hackers@postgresql.org>, Pavel Trukhanov <pavel.trukhanov@gmail.com>
Date: 2024-03-25T16:35:27Z
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. Introduce squashing of constant lists in query jumbling

  2. Make documentation builds reproducible

  3. Include values of A_Const nodes in query jumbling

  4. Teach planner about more monotonic window functions

  5. Split up guc.c for better build speed and ease of maintenance.

> On Sun, Mar 24, 2024 at 11:36:38PM +0900, Yasuo Honda wrote:
> Thanks for the information. I can apply these 4 patches from
> 0eb23285a2 . I tested this branch from Ruby on Rails and it gets some
> unexpected behavior from my point of view.
> Setting pg_stat_statements.query_id_const_merge_threshold = 5 does not
> normalize sql queries whose number of in clauses exceeds 5.
>
> Here are test steps.
> https://gist.github.com/yahonda/825ffccc4dcb58aa60e12ce33d25cd45#expected-behavior
>
> It would be appreciated if I can get my understanding correct.

From what I understand out of the description this ruby script uses
prepared statements, passing values as parameters, right? Unfortunately
the current version of the patch doesn't handle that, it works with
constants only [1]. The original incarnation of this feature was able to
handle that, but the implementation was considered to be not suitable --
thus, to make some progress, it was left outside.

The plan is, if everything goes fine at some point, to do a follow-up
patch to handle Params and the rest.

[1]: https://www.postgresql.org/message-id/20230211104707.grsicemegr7d3mgh%40erthalion.local