From mmahoui@lucy.cs.waikato.ac.nz Mon Apr  7 17:43 EDT 1997
Return-Path: <mmahoui@lucy.cs.waikato.ac.nz>
Received: from shasha.cs.nyu.edu by griffin.cs.nyu.edu (SMI-8.6/1.20)
	id RAA09181; Mon, 7 Apr 1997 17:43:01 -0400
Received: from cs.waikato.ac.nz by shasha.cs.nyu.edu (SMI-8.6/1.20)
	id RAA07696; Mon, 7 Apr 1997 17:43:16 -0400
Received: from ilias by cs.waikato.ac.nz (SMI-8.6/SMI-SVR4)
	id JAA29855; Tue, 8 Apr 1997 09:42:33 +1200
Message-Id: <1.5.4.32.19970407214126.0090d9e8@pop.cs.waikato.ac.nz>
X-Sender: mmahoui@pop.cs.waikato.ac.nz
X-Mailer: Windows Eudora Light Version 1.5.4 (32)
Mime-Version: 1.0
Date: Tue, 08 Apr 1997 09:41:26 +1200
To: shasha@shasha.cs.nyu.edu (Dennis Shasha)
From: Malika Mahoui <mmahoui@lucy.cs.waikato.ac.nz>
Subject: Re: ps copy of "The Performance of Current B...
Content-Type: text/plain; charset="us-ascii"
Content-Length: 793
Status: R

Dear Dennis,

Thanks alot for your quick answer.

Best regards,

Malika

PS: I was present in Montpellier when you presented your seminar in LSI
laboratory, 
and I really appreciated your French speaking.




At 17:30 7/04/97 -0400, you wrote:
>
>For the Tods paper, I have to ask Ted to send it to you.
>Ted, could you please do so.
>Thanks,
>Dennis
>
      ----------------------------------------------------------------------
---------
    /  Malika Mahoui-Guerni                   email:
mmahoui@cs.waikato.ac.nz       /
   /  Department of Computer Science          phone: +64-7-838-4406
/
  /  University of Waikato    		      fax:   +64-7-838-4155               /
 /  Private Bag 3105, Hamilton NZ
/
 ---------------------------------------------------------------------------
----- 


From pradipm@spectralogic.com Wed Nov 12 19:19 EST 1997
Return-Path: <pradipm@spectralogic.com>
Received: from cs.nyu.edu by griffin.cs.nyu.edu (SMI-8.6/1.20)
	id TAA20135; Wed, 12 Nov 1997 19:19:25 -0500
Received: from wali.com by cs.nyu.edu (SMI-8.6/1.20)
	id TAA07185; Wed, 12 Nov 1997 19:19:25 -0500
Received: from reactor.spectralogic.com (reactor [205.184.208.138]) by wali.com (8.6.12/8.6.12) with ESMTP id RAA21612 for <Shasha@Cs.Nyu.Edu>; Wed, 12 Nov 1997 17:20:21 -0700
Received: by reactor with Internet Mail Service (5.0.1457.3)
	id <WG5M8AR4>; Wed, 12 Nov 1997 17:18:39 -0700
Message-ID: <A14D973480EED011801000805FEAD7360ED98E@reactor>
From: PradipM <pradipm@spectralogic.com>
To: "'Prof. Dennis E. Shasha'" <Shasha@cs.nyu.edu>
Subject: Request Copy of Paper.
Date: Wed, 12 Nov 1997 17:18:39 -0700
X-Priority: 3
X-Mailer: Internet Mail Service (5.0.1457.3)
Content-Type: text
Content-Length: 836
Status: R

Hello Prof. Shasha,
I recently came across a reference to your VLDB'94 paper
with Theodore Johnson: "2Q: A Low Overhead High Performance
Buffer Management Replacement Algorithm, Pg. 439-450".

I live in Denver/Boulder area and work for a mass storage
company and naturally I searched the nearby Univ. libraries
for that VLDB'94.  Came up empty handed! Closest source
is Lawrence-Berkely at CA. (Fort Collins library got flooded
lately!)

I was wondering if you had a copy of that work (paper or
magnetic media), could you please send me one?  Me and
my company believe in lot of reading before conceiving or
implementing any kind of a cutting edge product.

With regards.

Pradip Maitra
Spectra Logic Corp.
1700 N, 55 St.,
Boulder, CO 80301.

Tel.        303-449-6444, x316,
FAX       303-939-8844.
E-mail    PradipM@SpectraLogic.Com



From alberto.lerner@gmail.com  Fri Dec  4 08:22:17 2009
Received: from mx2.cims.nyu.edu (mx2.cims.nyu.edu [128.122.81.61])
	by mail.cims.nyu.edu (8.13.8+Sun/8.13.8) with ESMTP id nB4DMGNI013661
	for <shasha@mail.cims.nyu.edu>; Fri, 4 Dec 2009 08:22:16 -0500 (EST)
Received: from mail-yw0-f182.google.com (mail-yw0-f182.google.com [209.85.211.182])
	by mx2.cims.nyu.edu (8.13.8+Sun/8.13.8) with ESMTP id nB4DMFsD005833
	for <shasha@courant.nyu.edu>; Fri, 4 Dec 2009 08:22:16 -0500 (EST)
Received: by ywh12 with SMTP id 12so2554832ywh.21
        for <shasha@courant.nyu.edu>; Fri, 04 Dec 2009 05:22:10 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=gmail.com; s=gamma;
        h=domainkey-signature:mime-version:received:in-reply-to:references
         :date:message-id:subject:from:to:content-type;
        bh=klVZnDCEwV/866dzEpjQ3Ok9SgdLLu95DEBIZtoCMWk=;
        b=ngZbnHi1DAh8EmWod+oiJX3GlPG/CjBze3zcypiRaO+0lzDpJyM1Eh7LoO0uTFs/rE
         H8O10iuR3ypTNo764eVZ5VOi3/1q3qjv7oLZxx2JDFu4JzBKUF35Y065UE5JHKavKnAB
         imEnpDNa7kpPEzZhar2x8+hjhgnYzEB1bDojM=
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type;
        b=FaVpY9Ph0I7eaeRJsBI8aJZWMLW+5Czp9Y3qBwwfosr5dc1rFFV2n0/+kyO6aMZRba
         7/+Tc0saeqrMM6GSxw2Ma9hi+DZhKXmrgB35PikHm7c7r3RGjob7E4ev/v+5H1RE1eLk
         yITwZOVJNQ1NovoCyX4OcIuY7yPUC3nvjfMP4=
MIME-Version: 1.0
Received: by 10.150.44.27 with SMTP id r27mr5347647ybr.263.1259932927818; Fri, 
	04 Dec 2009 05:22:07 -0800 (PST)
In-Reply-To: <200912040229.nB42Trq0017526@access2.cims.nyu.edu>
References: <200912040229.nB42Trq0017526@access2.cims.nyu.edu>
Date: Fri, 4 Dec 2009 08:22:07 -0500
Message-ID: <106a68620912040522g4da03f99tead1b60e67f34b65@mail.gmail.com>
Subject: Re: Concurrent buffer management?
From: Alberto Lerner <alberto.lerner@gmail.com>
To: Dennis Shasha <shasha@courant.nyu.edu>
Content-Type: text/plain; charset=ISO-8859-1
X-Scanned-By: MIMEDefang 2.58 on 128.122.81.61
X-Greylist: Sender DNS name whitelisted, not delayed by milter-greylist-3.0 (mx2.cims.nyu.edu [128.122.81.61]); Fri, 04 Dec 2009 08:22:16 -0500 (EST)
Status: R
Content-Length: 894

Dear Dennis,

Yes, the comparison article is in IEEE Computer. Ther ARC algorithm
writers made it. You got the sophistication part right... ;)

Hashing opens interesting possibilities, too. Thanks for the suggestion.

Best,
Alberto.

On Thu, Dec 3, 2009 at 9:29 PM, Dennis Shasha <shasha@courant.nyu.edu> wrote:
>
> Dear Alberto,
>
> 1. I'm honored.
>
> 2. It seems to me that you could handle shared pages using
> some kind of hashing scheme whereby there could be several 2q or 2q_alberto
> variants and that one chooses a queue at random to get a page but
> one could hash a page id to figure out which queue to release a page to.
>
> 3. I don't know this area any more but David Mazieres told me once
> that there was a comparison of a bunch of buffer algorithms including
> 2q. So that paper must have compared 2q with more sophisticated ideas
> (like the authors':).
>
> Best,
> Dennis
>

From alberto.lerner@gmail.com  Fri Dec  4 10:06:47 2009
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.80.107])
	by mail.cims.nyu.edu (8.13.8+Sun/8.13.8) with ESMTP id nB4F6lMP015587
	for <shasha@mail.cims.nyu.edu>; Fri, 4 Dec 2009 10:06:47 -0500 (EST)
Received: from mail-yw0-f183.google.com (mail-yw0-f183.google.com [209.85.211.183])
	by mx.cims.nyu.edu (8.13.8+Sun/8.13.8) with ESMTP id nB4F6j9G019350
	for <shasha@courant.nyu.edu>; Fri, 4 Dec 2009 10:06:45 -0500 (EST)
Received: by ywh13 with SMTP id 13so2583988ywh.29
        for <shasha@courant.nyu.edu>; Fri, 04 Dec 2009 07:06:40 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=gmail.com; s=gamma;
        h=domainkey-signature:mime-version:received:in-reply-to:references
         :date:message-id:subject:from:to:content-type;
        bh=zSuko1el+JS3aAAFpWVevNFzovJszIWBvueR/LMptuU=;
        b=tQy9GZatdzPw7M7pIiLxrECyuHqaBn4WtfEv8V3kcKO8uF+3nEG7/y9WO3KCEyLEef
         wl9el+YD0VOnftfJPc5ZhYtSDHb2hXFq34IlSUnJSdnK7jC729K5YPnXCA0fdthrC1qb
         PVq/OFsFhZC0dhvY3PU12Ab6zVbKOM5HZvvtc=
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type;
        b=YdMmSFQom3cD06B4701ZbLWrX0GLP4lrRzR1i72KCEpKuVa/uKUnKWehndL7DFZeUX
         caENd3fNBLMPqkKhGwdv7EDZYOvC3GERYP3JvJ4chKhSXWGCyIoY6Ue6ZmPRTLefsf4e
         MYXC92kot/TqPS3ORQSCmiummnAGnHWElDo0M=
MIME-Version: 1.0
Received: by 10.150.142.13 with SMTP id p13mr5682076ybd.25.1259939197439; Fri, 
	04 Dec 2009 07:06:37 -0800 (PST)
In-Reply-To: <200912041454.nB4EsZkc009976@access2.cims.nyu.edu>
References: <200912041454.nB4EsZkc009976@access2.cims.nyu.edu>
Date: Fri, 4 Dec 2009 10:06:37 -0500
Message-ID: <106a68620912040706y5607db69n5e7b611f7453a86@mail.gmail.com>
Subject: Re: Concurrent buffer management?
From: Alberto Lerner <alberto.lerner@gmail.com>
To: Dennis Shasha <shasha@courant.nyu.edu>
Content-Type: text/plain; charset=ISO-8859-1
X-Scanned-By: MIMEDefang 2.58 on 128.122.80.107
X-Greylist: Sender DNS name whitelisted, not delayed by milter-greylist-3.0 (mx.cims.nyu.edu [128.122.80.107]); Fri, 04 Dec 2009 10:06:47 -0500 (EST)
Status: R
Content-Length: 504

Sure, here is it, incidentally
http://theory.stanford.edu/~megiddo/pdf/IEEE_COMPUTER_0404.pdf
The results are really hair splitting. Although I met and respect Nimrod.

I'm still frustrated at how not intuitive is to make a simple cache
core-scalable... Fundamentally, maintaining any order among pages is
bottleneck.

Alberto.

On Fri, Dec 4, 2009 at 9:54 AM, Dennis Shasha <shasha@courant.nyu.edu> wrote:
>
> Alberto,
> Thanks for finding the article with such a meager description.
> Best,
> Dennis
>

From queue@etri.re.kr  Mon Mar  5 19:40:19 2012
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.14.4+Sun/8.14.4) with ESMTP id q260eI6w001991
	for <shasha@mail.cims.nyu.edu>; Mon, 5 Mar 2012 19:40:18 -0500 (EST)
Received: from email2.etri.info (email2.etri.re.kr [129.254.16.132])
	by mx.cims.nyu.edu (8.14.4+Sun/8.14.4) with ESMTP id q260eGRd001206
	for <shasha@courant.nyu.edu>; Mon, 5 Mar 2012 19:40:16 -0500 (EST)
Received: from cyberlation.etri.re.kr ([129.254.191.22]) by email2.etri.info with Microsoft SMTPSVC(6.0.3790.3959);
	 Tue, 6 Mar 2012 09:17:16 +0900
Subject: Re: Questions on 2Q algorithm presented in 1994
Mime-Version: 1.0 (Apple Message framework v1257)
Content-Type: text/plain; charset=euc-kr
From: =?euc-kr?B?udrBvrHZ?= <queue@etri.re.kr>
In-Reply-To: <201203051157.q25Bv0qZ027615@access2.cims.nyu.edu>
Date: Tue, 6 Mar 2012 09:17:16 +0900
Cc: shasha@cs.nyu.edu, ted@cise.ufl.edu
Content-Transfer-Encoding: quoted-printable
Message-Id: <FAFB2A61-E31D-4D72-852C-0796D83F6262@etri.re.kr>
References: <201203051157.q25Bv0qZ027615@access2.cims.nyu.edu>
To: Dennis Shasha <shasha@courant.nyu.edu>
X-Mailer: Apple Mail (2.1257)
X-OriginalArrivalTime: 06 Mar 2012 00:17:16.0802 (UTC) FILETIME=[7CD0AE20:01CCFB2E]
X-Scanned-By: MIMEDefang 2.71 on 128.122.49.99
X-Greylist: Delayed for 00:22:53 by milter-greylist-4.2.6 (mx.cims.nyu.edu [128.122.49.99]); Mon, 05 Mar 2012 19:40:18 -0500 (EST)
Status: R
Content-Length: 1715

Dear Dr. Shasha,
Thank you for your reply.
Your answer is really helpful to me.
But I cannot reach to Dr. T. Johnson since the email address is =
undeliverable.

Have a good day!
Best regards,

Jong-Geun Park


2012. 3. 5., =BF=C0=C8=C4 8:57, Dennis Shasha =C0=DB=BC=BA:

>=20
>=20
> *** Dear Colleague, I will attempt an answer, but Ted may remember =
better.
>=20
>=20
> =46rom queue@etri.re.kr  Mon Mar  5 03:48:16 2012
>=20
> Dear, Dr. T. Johnson and D. Shasha
> I am a Korean (Republic of) engineer interested in the cache =
replacement =3D
> algorithms and now review remarkable previous studies including your =
2Q =3D
> algorithm.
> I am not sure whether you can answer my following questions now, since =
=3D
> it was presented in 1994.
> But I hope you can give me answers.
>=20
> Q1. In 2Q Full version algorithm, why a page X is NOT moved to the =
head =3D
> of Am queue when X is in A1in queue.
> =3D46rom the algorithm, when it does, nothing to be done.
>=20
> *** Because it may be a correlated reference, i.e. a reference that
> occurs immediately after another one by the same thread, but not
> one that indicates lasting importance.
>=20
>=20
> Q2, When you do trace-driven simulations with DB2 logs, which =
algorithm =3D
> was applied between simplified an full version of 2Q ?
>=20
> *** full.
>=20
> Q3, Do you have DB2 logs which were experimented in your paper, and =3D
> could you send me them ?
>=20
> *** I do not. Ted may. Best, Dennis
>=20
> I hope to experiment to compare the performance of algorithms among =
2Q, =3D
> my proposed algorithm, and others.
>=20
> If possible, I sincerely wish to get answers from you.
> Thank you so much.
>=20
> Sincerely yours,
> Jong-Geun Park
>=20

