Hello Andres and Jeff,
18.09.2026 21:08, Jeff Davis wrote:
On Fri, 2026-09-18 at 12:56 -0400, Andres Freund wrote:
To do better, we would need to use a
* self-exclusive lock, perhaps ShareUpdateExclusiveLock, here and
before
* *every* CatalogTupleUpdate() of a row that GRANT/REVOKE can
affect.
...
I also am not convinced that cancelling autovacs in response to a
command like
this would be the end of the world.
For GRANT/REVOKE, that matches my intuition. But if we need a stronger
lock before every modification of a catalog row, that's a larger change
that would affect lots of DDL.
I've recalled a known issue (still reproduced on current master) with
VACUUM and GRANT locks, which might be interesting to you in this context:
https://www.postgresql.org/message-id/9f7cc148-55d1-4062-8229-b6886d7aa380%40gmail.com
Best regards,
Alexander
