Feedback-ID: ib7d648f3:Fastmail
MIME-Version: 1.0
Date: Thu, 05 Jun 2025 21:55:04 +0330
From: "Mahdi Bahrami" <pgsql@mahdibm.com>
To: "Mahdi Bahrami" <pgsql@mahdibm.com>,
 "Thomas Munro" <thomas.munro@gmail.com>
Cc: =?UTF-8?Q?=C3=81lvaro_Herrera?= <alvherre@kurilemu.de>,
 pgsql-performance@lists.postgresql.org
Message-Id: <2ba6b290-8322-4c48-a6a1-65409220d1c3@app.fastmail.com>
In-Reply-To: <ce543641-8940-4eec-88c4-3adc498f3e2d@app.fastmail.com>
References: <202505271507.xvfbdztae74a@alvherre.pgsql>
 <ad74d853-bd28-4e72-9ba5-1ca624c2d879@app.fastmail.com>
 <CA+hUKGJaCTTspp7bQPG2nwDmzk_XzAdNvFwcx4sWHt2QUEK4QA@mail.gmail.com>
 <ce543641-8940-4eec-88c4-3adc498f3e2d@app.fastmail.com>
Subject: Re: Database creation performance drop going from pg 14 to pg 15+
Content-Type: multipart/alternative;
 boundary=4a2de8a4d9014f84897081244443695e
Archived-At: 
 <https://www.postgresql.org/message-id/2ba6b290-8322-4c48-a6a1-65409220d1c3%40app.fastmail.com>
Precedence: bulk

--4a2de8a4d9014f84897081244443695e
Content-Type: text/plain
Content-Transfer-Encoding: 7bit

Here's what the OP of the PostgresNIO issue has mentioned about what performance impact usage of `file_copy` has in his setup (https://github.com/SwiftPackageIndex/SwiftPackageIndex-Server/pull/3812):

`Series: Default Strategy
Suite AllTests passed after 5.081 seconds
Suite AllTests passed after 5.274 seconds
Suite AllTests passed after 5.306 seconds
Suite AllTests passed after 5.224 seconds
Suite AllTests passed after 6.343 seconds
Suite AllTests passed after 5.450 seconds
Series: file_copy Strategy
Suite AllTests passed after 4.729 seconds
Suite AllTests passed after 4.755 seconds
Suite AllTests passed after 4.739 seconds
Suite AllTests passed after 4.772 seconds
Suite AllTests passed after 4.866 seconds
Suite AllTests passed after 4.855 seconds`

This is them running the tests in parallel multiple times with and without file_copy, just to benchmark.
Overall ~0.5-6s / ~10% speedup, which isn't too significant at this level, but is still a noticeable relative improvement.

--4a2de8a4d9014f84897081244443695e
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title></head><body><div>Here's what =
the OP of the PostgresNIO issue has mentioned about what performance imp=
act usage of `file_copy` has in his setup (<a href=3D"https://github.com=
/SwiftPackageIndex/SwiftPackageIndex-Server/pull/3812">https://github.co=
m/SwiftPackageIndex/SwiftPackageIndex-Server/pull/3812</a>):</div><div><=
br></div><pre style=3D"border-top-width:1px;border-right-width:1px;borde=
r-bottom-width:1px;border-left-width:1px;border-top-style:solid;border-r=
ight-style:solid;border-bottom-style:solid;border-left-style:solid;borde=
r-top-color:rgb(204, 204, 204);border-right-color:rgb(204, 204, 204);bor=
der-bottom-color:rgb(204, 204, 204);border-left-color:rgb(204, 204, 204)=
;border-image-source:initial;border-image-slice:initial;border-image-wid=
th:initial;border-image-outset:initial;border-image-repeat:initial;borde=
r-top-left-radius:3px;border-top-right-radius:3px;border-bottom-right-ra=
dius:3px;border-bottom-left-radius:3px;background-image:initial;backgrou=
nd-position-x:initial;background-position-y:initial;background-size:init=
ial;background-repeat:initial;background-attachment:initial;background-o=
rigin:initial;background-clip:initial;background-color:rgb(246, 246, 246=
);font-family:menlo, consolas, monospace;font-size:90%;margin-top:7px;ma=
rgin-right:0px;margin-bottom:7px;margin-left:0px;padding-top:7px;padding=
-right:10px;padding-bottom:7px;padding-left:10px;white-space-collapse:pr=
eserve;text-wrap-mode:wrap;overflow-wrap:break-word;"><code style=3D"bor=
der-top-width:1px;border-right-width:1px;border-bottom-width:1px;border-=
left-width:1px;border-top-style:solid;border-right-style:solid;border-bo=
ttom-style:solid;border-left-style:solid;border-top-color:rgb(204, 204, =
204);border-right-color:rgb(204, 204, 204);border-bottom-color:rgb(204, =
204, 204);border-left-color:rgb(204, 204, 204);border-image-source:initi=
al;border-image-slice:initial;border-image-width:initial;border-image-ou=
tset:initial;border-image-repeat:initial;border-top-left-radius:3px;bord=
er-top-right-radius:3px;border-bottom-right-radius:3px;border-bottom-lef=
t-radius:3px;background-image:initial;background-position-x:initial;back=
ground-position-y:initial;background-size:initial;background-repeat:init=
ial;background-attachment:initial;background-origin:initial;background-c=
lip:initial;background-color:rgb(246, 246, 246);font-family:menlo, conso=
las, monospace;font-size:90%;padding-top:1px;padding-right:3px;padding-b=
ottom:1px;padding-left:3px;">Series: Default Strategy=0ASuite AllTests p=
assed after 5.081 seconds=0ASuite AllTests passed after 5.274 seconds=0A=
Suite AllTests passed after 5.306 seconds=0ASuite AllTests passed after =
5.224 seconds=0ASuite AllTests passed after 6.343 seconds=0ASuite AllTes=
ts passed after 5.450 seconds=0ASeries: file_copy Strategy=0ASuite AllTe=
sts passed after 4.729 seconds=0ASuite AllTests passed after 4.755 secon=
ds=0ASuite AllTests passed after 4.739 seconds=0ASuite AllTests passed a=
fter 4.772 seconds=0ASuite AllTests passed after 4.866 seconds=0ASuite A=
llTests passed after 4.855 seconds</code></pre><div><br>This is them run=
ning the tests in parallel multiple times with and without file_copy, ju=
st to benchmark.<br>Overall ~0.5-6s / ~10% speedup, which isn't too sign=
ificant at this level, but is still a noticeable relative improvement.</=
div><div><br></div></body></html>
--4a2de8a4d9014f84897081244443695e--