Horton

Re: [PATCH] Proposal to Enable/Disable Index using ALTER INDEX

Sami Imseih <samimseih@gmail.com>

View thread
From: Sami Imseih <samimseih@gmail.com>
To: David Rowley <dgrowleyml@gmail.com>
Cc: Robert Treat <rob@xzilla.net>, Shayon Mukherjee <shayonj@gmail.com>, jian he <jian.universality@gmail.com>, Gurjeet Singh <gurjeet@singh.im>, Nathan Bossart <nathandbossart@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2025-07-18T02:25:17Z
Lists: pgsql-hackers 
> My concern with #1 is that when we one day do get query hints, we'll
> be left with a bunch of redundant ways to influence planner behaviour.


The GUC is more than just a hint. It can serve as a hint, but it also offers
operational benefits. For example, if I mark an index as invisible and that
only affects a subset of my workload, I don’t need to make the index visible
again. Instead, I can tune that specific workload to operate without it.

Once I’m confident the workload performs well, I can safely drop the index.

I’d argue we should not provide the ALTER option without the GUC, for
more granular control.
Regards,

Sami