Feedback-ID: id4a34324:Fastmail
Date: Tue, 24 Mar 2026 13:53:43 -0400
From: Andres Freund <andres@anarazel.de>
To: Melanie Plageman <melanieplageman@gmail.com>
Cc: Kirill Reshke <reshkekirill@gmail.com>,
	Chao Li <li.evan.chao@gmail.com>, Andrey Borodin <x4mmm@yandex-team.ru>,
	Xuneng Zhou <xunengzhou@gmail.com>, Robert Haas <robertmhaas@gmail.com>,
	PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>,
 Heikki Linnakangas <hlinnaka@iki.fi>
Subject: Re: eliminate xl_heap_visible to reduce WAL (and eventually set VM
 on-access)
Message-ID: <itvgqc6vncbjsjfmrptfvkkeg5vqzhalaguya2z77t6c6ctpc3@wsdrgbn4bxaa>
References: 
 <CAAKRu_Z8Ry_ynNBPAzs_Ry3MQi9NaBgt1ccLgwRsDbxWpocaBg@mail.gmail.com>
 <CAAKRu_ZbOp52rnkjf63h5mf94raEKBH7AAbz6QTx-xdH9yLfmQ@mail.gmail.com>
 <CAAKRu_b8m+iuupm4ZX+2_V5Xj5u4jCTrU=Tv=epg6p4H2SMkFQ@mail.gmail.com>
 <CALdSSPh9hVXNiPwdntWqbMzu5upKy0jBDDe7Un0_Nf2A54R2VQ@mail.gmail.com>
 <CAAKRu_a6Cd7JnxhY4A=b_Paxc3UDUDOPeqV3GbzMh=R2KkD-uQ@mail.gmail.com>
 <jlsvov4o5xswxjvjhvuchz6y55ncvoc457grvxct7cub5gcxuj@4e2toesujnpr>
 <CAAKRu_bsHbvt+VqbjHXsdphKf8fqwBEutRhH3fmo+qUVe=yBKw@mail.gmail.com>
 <CAAKRu_ZhHtEaucO--SdYrCjq0zgqk_LPztUD+-QS74A2htXgKw@mail.gmail.com>
 <CAAKRu_Zj8G4T=HN3QSY7iQvkKSQk-k1fq+eJkjCBNqoSg63z+Q@mail.gmail.com>
 <CAAKRu_bgP-DMZs=D2j2N0+U9+uWU5cVagw-yZLOuhYbWj_KwnA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CAAKRu_bgP-DMZs=D2j2N0+U9+uWU5cVagw-yZLOuhYbWj_KwnA@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/itvgqc6vncbjsjfmrptfvkkeg5vqzhalaguya2z77t6c6ctpc3%40wsdrgbn4bxaa>
Precedence: bulk

Hi,

On 2026-03-23 17:54:13 -0400, Melanie Plageman wrote:
> I've made some significant changes to 0006 and realized I need some
> help. 0006 tracks what relations are modified by a query. This new
> version (v44) uses relation oids instead of rt indexes to handle cases
> where the same relation appears more than once in the range table
> (e.g. INSERT INTO foo SELECT * FROM foo; foo appears twice). It
> computes modifiedRelOids (a list of relation OIDs modified by the
> query) in the planner and stores them in the PlannedStmt. There is one
> big issue I'm not sure how to solve:

I'm not entirely sure this is something we need to catch and therefore not
sure that modifiedRelOids is worth the trouble over just having the RT
indexes.


> For queries like INSERT INTO ptable SELECT * FROM ptable, where ptable
> is a partitioned table, though we scan ptable, we don't know when
> executing that scan that we will then modify ptable with the insert.

But does that matter? If such a query inserts a meaningful amount of rows it's
going to insert into different pages than the ones you selected from?


> In my patch, I've added find_all_inheritors() when populating
> modifiedRelOids, but I realize this probably isn't acceptable to add
> to planner from a performance perspective.

Agreed.


Greetings,

Andres Freund