From: | Peter Geoghegan <pg(at)bowt(dot)ie> |
---|---|
To: | Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> |
Cc: | PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
Subject: | Re: vacuum_cost_page_miss default value and modern hardware |
Date: | 2021-01-14 00:24:12 |
Message-ID: | CAH2-Wzk7wbWecbSAEq+0dgydR7E6-XVaFjXLF1_UhB+ThywQQw@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Wed, Jan 6, 2021 at 7:43 PM Peter Geoghegan <pg(at)bowt(dot)ie> wrote:
> More concretely, we could perhaps lower vacuum_cost_page_miss to 5. It
> has had the value as 10 as its default since 2004 (just like
> vacuum_cost_page_dirty, whose default has also not been changed since
> the start). These defaults were decided in a time when nbtree VACUUM
> could do lots of random I/O, there was no visibility map, etc. So this
> refresh is not just about hardware.
Attached patch lowers vacuum_cost_page_miss to 3. I think that this
change in the default is both likely to be helpful in medium to large
installations, and unlikely to cause harm in small installations. If
I/O for reads made by VACUUM is naturally very slow (even in the
common case where it's entirely sequential), then that will naturally
provide additional throttling.
I will add this to the final CF for Postgres 14.
Thanks
--
Peter Geoghegan
Attachment | Content-Type | Size |
---|---|---|
v1-0001-Lower-vacuum_cost_page_miss-s-default-to-3.patch | application/octet-stream | 2.5 KB |
From | Date | Subject | |
---|---|---|---|
Next Message | tsunakawa.takay@fujitsu.com | 2021-01-14 00:27:16 | RE: Disable WAL logging to speed up data loading |
Previous Message | Tom Lane | 2021-01-13 23:49:25 | Re: [DOC] Document concurrent index builds waiting on each other |