Received: from malur.postgresql.org ([2a02:16a8:dc51::56]) by arkaria.postgresql.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA384:256) (Exim 4.89) (envelope-from ) id 1fftlZ-0001bh-8h for pgsql-hackers@arkaria.postgresql.org; Wed, 18 Jul 2018 21:13:45 +0000 Received: from localhost ([127.0.0.1] helo=malur.postgresql.org) by malur.postgresql.org with esmtp (Exim 4.89) (envelope-from ) id 1fftlX-0000P1-Dy for pgsql-hackers@arkaria.postgresql.org; Wed, 18 Jul 2018 21:13:43 +0000 Received: from makus.postgresql.org ([2001:4800:1501:1::229]) by malur.postgresql.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA384:256) (Exim 4.89) (envelope-from ) id 1fftlX-0000Ot-17 for pgsql-hackers@lists.postgresql.org; Wed, 18 Jul 2018 21:13:43 +0000 Received: from mail-pg1-x543.google.com ([2607:f8b0:4864:20::543]) by makus.postgresql.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.89) (envelope-from ) id 1fftlU-0007p4-1M for pgsql-hackers@lists.postgresql.org; Wed, 18 Jul 2018 21:13:41 +0000 Received: by mail-pg1-x543.google.com with SMTP id n7-v6so2556918pgq.4 for ; Wed, 18 Jul 2018 14:13:39 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=message-id:mime-version:to:cc:from:subject:date:importance :in-reply-to:references; bh=uTCMThtXgbMeWI2uAQ44ghAJpDNlAsOlwWjYB19ev+I=; b=UcYlYXjWCKoXHMqMYIkHV55E4wTJ/gRGE6nAsMyUe/qkSvJpJElyhF4t7Y62TN5bP6 9mWO98QiKUhRbYpzDvUtW+adxCRkesW07NqoH6kV7iQ5DMZAQ0iXWC7CniEZ9QoKcGOH WJQAvSB9rcKEKfpt4EWScivsPJNC71lAWU72rvXTzwobsc+DlU29dRh9Vx+Bbm2HjL9T T8gbgTrBVWrfDms/BJ6trK8K3K/lOoF49KX30ihIOV3AZeBKmeJER7YpUXdc4Yn8KYgr fY6iylJ8yKt0G1VgHGvrntbluMRK1FakeKGM5DPwFIizmWtjeBmw35K10KiDEkan2ci9 RnmA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:message-id:mime-version:to:cc:from:subject:date :importance:in-reply-to:references; bh=uTCMThtXgbMeWI2uAQ44ghAJpDNlAsOlwWjYB19ev+I=; b=Nh6sbV/UcFOnJRkjsvr9gF11J80r7ZgvvXmfIscIWPXl1CutD4inmfWP6pF3qSM9Ye lYEwpb5pBBl4SNBG70FcmujPFiP1iAa29PwbCTO4SSxxU87jlHGX5ZXcvzvIbVJASUJU HwJPXy/T/qbrMFCpbTuka4ivzaWgu/Z5MIzugoe3rk/taY9WljLsHfysROAHZgf/SjoD 3ww2bzi/jjL+dl2X/4o9/XLfXnV6oe+F/UJ4bT+Z/OIk2Jc0+EzwqWsk6YftCUkMnjEs pJcOu0ZR6yvmMILHSuJLUeBQS9t9VJPXLWQ95hu+PkCm8J212RpSyfs2U2WM0ITS+vRA cLfQ== X-Gm-Message-State: AOUpUlHKuTeer3EGPgc15Zh+6jN/wVYuDzV307+OMKMOVpZcHMtlKogk N2kBlSLkCUbCCPv1msHPELE= X-Google-Smtp-Source: AAOMgpeJ9+4CuM4JW90Say+Sy1HyuAGW7dBvhzhl5y8mg9SSM345fhixIRkcrgkY2rcO0Jg4eODNbQ== X-Received: by 2002:a65:40ca:: with SMTP id u10-v6mr7123528pgp.2.1531948418880; Wed, 18 Jul 2018 14:13:38 -0700 (PDT) Received: from ?IPv6:::ffff:207.23.222.210? (d207-023-222-210.wireless.sfu.ca. [207.23.222.210]) by smtp.gmail.com with ESMTPSA id o27-v6sm14358071pfj.35.2018.07.18.14.13.37 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Wed, 18 Jul 2018 14:13:38 -0700 (PDT) Message-ID: <5b4fad82.1c69fb81.7a6f6.d32b@mx.google.com> MIME-Version: 1.0 To: Tomas Vondra Cc: Andrey Borodin , Peter Geoghegan , PostgreSQL Hackers From: Kefan Yang Subject: RE: GSOC 2018 Project - A New Sorting Routine Date: Wed, 18 Jul 2018 14:13:38 -0700 Importance: normal X-Priority: 3 In-Reply-To: <78b9c822-e127-5250-a01c-4c51a50c82bd@2ndquadrant.com> References: <57ae7691-6dc8-d6d0-361a-a7269e656273@2ndquadrant.com> <00147235-69d3-d4f0-36c1-198d630af7ad@2ndquadrant.com> <1C5B17F1-D37F-4730-A2E4-EB7288C6DC09@yandex-team.ru> <78b9c822-e127-5250-a01c-4c51a50c82bd@2ndquadrant.com> Content-Type: multipart/alternative; boundary="_D37BD020-810B-45CE-99F7-AC86736DD3E1_" List-Id: List-Help: List-Subscribe: List-Post: List-Owner: List-Archive: Precedence: bulk --_D37BD020-810B-45CE-99F7-AC86736DD3E1_ Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Hey Tomas! I am trying to reproduce the results on my machine. Could you please share = the script to generate .ods files? Regards, Kefan From: Tomas Vondra Sent: July 18, 2018 2:05 AM To: Andrey Borodin Cc: Peter Geoghegan; Kefan Yang; PostgreSQL Hackers Subject: Re: GSOC 2018 Project - A New Sorting Routine On 07/18/2018 07:06 AM, Andrey Borodin wrote: > Hi, Tomas! >=20 >> 15 =D0=B8=D1=8E=D0=BB=D1=8F 2018 =D0=B3., =D0=B2 1:20, Tomas Vondra > > =D0=BD=D0=B0=D0=BF=D0=B8=D1=81=D0= =B0=D0=BB(=D0=B0): >> >> So I doubt it's this, but I've tweaked the scripts to also set this GUC >> and restarted the tests on both machines. Let's see what that does. >=20 > Do you observe any different results? >=20 It did change the CREATE INDEX results, depending on the scale. The full=20 data is available at [1] and [2], attached is a spreadsheet summary from=20 the Xeon box. For the largest scale (1M rows) the regressions for CREATE INDEX queries=20 mostly disappeared. For 10k rows it still affects CREATE INDEX with a=20 text column, and the 100k case behaves just like before (so significant=20 regressions for CREATE INDEX). I don't have time to investigate this further at the moment, but I'm=20 still of the opinion that there's little to gain by replacing our=20 current sort algorithm with this. [1] https://bitbucket.org/tvondra/sort-intro-sort-xeon/src/master/ [2] https://bitbucket.org/tvondra/sort-intro-sort-i5/src/master/ regards --=20 Tomas Vondra http://www.2ndQuadrant.com PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services --_D37BD020-810B-45CE-99F7-AC86736DD3E1_ Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset="utf-8"

Hey Tomas!

<= o:p> 

I am trying to reproduce the resul= ts on my machine. Could you please share the script to generate .ods files?=

 

Regards,=

Kefan

 

From: Tomas Vondra
Sent: July 18, 2018 2:05 AM
To: Andrey Borodin
Cc: Peter Geoghegan; Kefan Yang; PostgreSQL Hackers
Subject: Re: GSOC 2018 Project - A = New Sorting Routine

 

 

 

On 07/18/2018 07:06 AM, Andrey Borodin wrote:

=

> Hi, Tomas!

>

>> 15 =D0=B8=D1=8E=D0=BB=D1=8F 2018 =D0=B3., =D0=B2 1= :20, Tomas Vondra <tomas.vondra@2ndquadrant.com

>> <mailto:tomas.vondra@2ndquadrant.com>> =D0=BD=D0=B0=D0= =BF=D0=B8=D1=81=D0=B0=D0=BB(=D0=B0):

>>&= nbsp;

>> So I doubt it's this, but I've= tweaked the scripts to also set this GUC

>> = and restarted the tests on both machines. Let's see what that does.

>

> Do you observe any diff= erent results?

>

&= nbsp;

It did change the CREATE INDEX results,= depending on the scale. The full

data is availabl= e at [1] and [2], attached is a spreadsheet summary from

the Xeon box.

 

For the largest scale (1M rows) the regressions for CREATE IND= EX queries

mostly disappeared. For 10k rows it sti= ll affects CREATE INDEX with a

text column, and th= e 100k case behaves just like before (so significant

regressions for CREATE INDEX).

 =

I don't have time to investigate this further at t= he moment, but I'm

still of the opinion that there= 's little to gain by replacing our

current sort al= gorithm with this.

 

 

[1] https://bitbucket.o= rg/tvondra/sort-intro-sort-xeon/src/master/

[2] htt= ps://bitbucket.org/tvondra/sort-intro-sort-i5/src/master/

 

regards

 

--

Tomas Vondra=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 http://www.2ndQuadrant.com

PostgreSQL Development, 24x7 Support, Remote DBA, Traini= ng & Services

 

= --_D37BD020-810B-45CE-99F7-AC86736DD3E1_--