Re: effective_io_concurrency's steampunk spindle maths
Thomas Munro <thomas.munro@gmail.com>
From: Thomas Munro <thomas.munro@gmail.com>
To: Tomas Vondra <tomas.vondra@2ndquadrant.com>
Cc: Andres Freund <andres@anarazel.de>,
pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2020-03-06T21:33:03Z
Lists: pgsql-hackers
On Sat, Mar 7, 2020 at 8:35 AM Tomas Vondra <tomas.vondra@2ndquadrant.com> wrote: > I think the main issue with keeping the current GUC name is that if you > had a value that worked, we'll silently interpret it differently. Which > is a bit annoying :-( Yeah. Perhaps we should just give the formula for translating v12 settings to v13 settings in the release notes. If we don't rename the GUC, you won't be forced to contemplate this when you upgrade, so the amount of prefetching we do will go down a bit given the same value. That is indeed what led me to start thinking about what a good new name would be. Now that I've been talked out of the "random_page" part, your names look like sensible candidates, but I wonder if there is some way to capture that it's "per operation"...
Commits
-
Introduce a maintenance_io_concurrency setting.
- fc34b0d9de27 13.0 landed
-
Simplify the effective_io_concurrency setting.
- b09ff53667ff 13.0 landed