Re: Pre-allocating WAL files

From: "Bossart, Nathan" <bossartn(at)amazon(dot)com>
To: Maxim Orlov <m(dot)orlov(at)postgrespro(dot)ru>, "pgsql-hackers(at)lists(dot)postgresql(dot)org" <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Pre-allocating WAL files
Date: 2021-11-10 18:59:43
Message-ID: CB15BEBD-98FC-4E72-86AE-513D34014176@amazon.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 10/8/21, 1:55 PM, "Bossart, Nathan" <bossartn(at)amazon(dot)com> wrote:
> Here is a first attempt at adding the pre-allocation logic to the
> checkpointer. I went ahead and just used CheckpointWriteDelay() for
> pre-allocating during checkpoints. I've done a few pgbench runs, and
> it seems to work pretty well. Initialization is around 15% faster,
> and I'm seeing about a 5% increase in TPS with a simple-update
> workload with wal_recycle turned off. Of course, these improvements
> go away once segments can be recycled.

Here is a rebased version of this patch set. I'm getting the sense
that there isn't a whole lot of interest in this feature, so I'll
likely withdraw it if it goes too much longer without traction.

Nathan

Attachment Content-Type Size
v7-0001-Move-WAL-segment-creation-logic-to-its-own-functi.patch application/octet-stream 7.5 KB
v7-0002-WAL-segment-pre-allocation.patch application/octet-stream 24.0 KB

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Jonathan S. Katz 2021-11-10 19:07:43 Re: 2021-11-11 release announcement draft
Previous Message Bossart, Nathan 2021-11-10 18:52:04 Re: archive modules