Re: [PATCH] Proposal to Enable/Disable Index using ALTER INDEX
Julien Rouhaud <rjuju123@gmail.com>
From: Julien Rouhaud <rjuju123@gmail.com>
To: Michael Paquier <michael@paquier.xyz>
Cc: Shayon Mukherjee <shayonj@gmail.com>, Sami Imseih <samimseih@gmail.com>, Robert Treat <rob@xzilla.net>, David Rowley <dgrowleyml@gmail.com>, jian he <jian.universality@gmail.com>, Gurjeet Singh <gurjeet@singh.im>, Nathan Bossart <nathandbossart@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2025-08-04T04:34:14Z
Lists: pgsql-hackers
On Mon, Aug 04, 2025 at 01:19:46PM +0900, Michael Paquier wrote: > On Sat, Aug 02, 2025 at 03:09:19PM -0400, Shayon Mukherjee wrote: > > Per this plan, this would mean that pg_hint_plan would need to > > support index-level hints, and it’s not a massive / impossible > > task. But it also means that both systems aren't fighting/overriding > > each other or making it difficult for users to understand when > > exactly an index is being used or not. Ultimately, this would also > > mean that `ALTER INDEX INVISIBLE` is a one-way door, and there is > > only one way to control index visibility in core, which makes sense > > to me. > > Sami has proposed a patch for pg_hint_plan that goes in the direction > of a DisableIndex hint, which is doable even if it requires some work > inside the extension code like some refactoring: > https://github.com/ossc-db/pg_hint_plan/issues/226 > > So I cannot say much about this proposal for core, but for > pg_hint_plan I'm looking at adding that for the v18 release of the > module, which is planned for the end of August/beginning of September. FWIW if anyone actually needs it, that feature had existed for decades in plantuner: http://www.sigaev.ru/git/gitweb.cgi?p=plantuner.git;a=summary