From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtp121.iad3a.emailsrvr.com (smtp121.iad3a.emailsrvr.com [173.203.187.121]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by huchra.bufferbloat.net (Postfix) with ESMTPS id 4F14921F13F for ; Fri, 20 Dec 2013 13:22:31 -0800 (PST) Received: from localhost (localhost.localdomain [127.0.0.1]) by smtp24.relay.iad3a.emailsrvr.com (SMTP Server) with ESMTP id 2C3751500C9; Fri, 20 Dec 2013 16:22:30 -0500 (EST) X-Virus-Scanned: OK Received: from app28.wa-webapps.iad3a (relay.iad3a.rsapps.net [172.27.255.110]) by smtp24.relay.iad3a.emailsrvr.com (SMTP Server) with ESMTP id 10B471500B8; Fri, 20 Dec 2013 16:22:30 -0500 (EST) Received: from reed.com (localhost.localdomain [127.0.0.1]) by app28.wa-webapps.iad3a (Postfix) with ESMTP id 0257A803A7; Fri, 20 Dec 2013 16:22:30 -0500 (EST) Received: by apps.rackspace.com (Authenticated sender: dpreed@reed.com, from: dpreed@reed.com) with HTTP; Fri, 20 Dec 2013 16:22:30 -0500 (EST) Date: Fri, 20 Dec 2013 16:22:30 -0500 (EST) From: dpreed@reed.com To: "Rich Brown" MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_20131220162230000000_26721" Importance: Normal X-Priority: 3 (Normal) X-Type: html In-Reply-To: <1D691810-3DD1-4899-9F63-E1713BD01840@gmail.com> References: <1D691810-3DD1-4899-9F63-E1713BD01840@gmail.com> Message-ID: <1387574550.00762811@apps.rackspace.com> X-Mailer: webmail7.0 Cc: "cerowrt-devel@lists.bufferbloat.net" Subject: Re: [Cerowrt-devel] Anything but "AQM" X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: Development issues regarding the cerowrt test router project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 20 Dec 2013 21:22:32 -0000 ------=_20131220162230000000_26721 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable =0AGiven that there is no likelihood of making localized queue management "= intelligent" because it has no global information whatsoever, I strongly su= ggest that "smart" "intelligent" and even "active" are hugely misleading.= =0A =0AThey are based on a completely false premise - that queues should be= allowed to build at all, and that local information can solve highly trans= ient global problems.=0A =0A"Dumb Queue Management" is going to be far supe= rior. Keep the queue at zero length, and try to be fair.=0A =0AThere's a s= imple way to do the latter - use a filter (similar to a Bloom filter) that = captures recent/frequent users of the queue, and when the queue on an outbo= und link grows more than about 2-3 packets (double buffering is all you nee= d to keep the link full) discard the most recent and frequent packets (or s= end information that tells them to slow down).=0A =0AThere's been a lot of = wasted time and effort trying to build queues long enough so that you can b= e "intelligent", but by then you have already lost the battle. You've gott= en into a positive feedback loop where you have encouraged the endpoints to= send more packets than you can ever drain out of the queue.=0A =0AI truly,= truly do not understand why people don't look at realistic network loads a= nd structures.=0A =0A=0A=0AOn Friday, December 20, 2013 3:52pm, "Rich Brown= " said:=0A=0A=0A=0A> Dave,=0A> =0A> You wrote:=0A= > =0A> > What's in a name? AQM has been pretty thoroughly defined to equal= =0A> > active queue *length* management and not packet scheduling.=0A> > Ov= erloading "AQM" what cerowrt does is apt to cause even more=0A> > confusion= in the field than it already does. We discussed using LBO as=0A> > a word = but that appears hopelessly overloaded with leveraged buy out.=0A> >=0A> > = I go back to one I liked a while back:=0A> >=0A> > Smart Queue Management. = (SQM)=0A> >=0A> > This got dissed on the aqm list too, but so far a viable = alternative=0A> > TLA has not appeared. It's sufficiently different to hang= a different=0A> > definition off of ("Smart queue management is an intelli= gent=0A> > combination of better packet scheduling (flow queuing) technique= s=0A> > along with with active queue length management (aqm)=E2=80=9D)=0A> = =0A> and=0A> =0A> > Any ideas for a name for packet scheduling, prioritizat= ion, and active=0A> > queue management better than just "AQM", or "QoS"?=0A= > >=0A> > SQM "Smarter Queue Management"=0A> > CeroShaper=0A> > LBO Latency= and Bandwidth Optimisation=0A> =0A> I was prepared to agree with =E2=80=9C= SQM=E2=80=9D, and had written a long note=0A> (below) when my brain uttered= =E2=80=9CIntelligent Queue Management=E2=80=9D.=0A> I=E2=80=99m not convin= ced that one is better than the other=E2=80=A6=0A> =0A> Rich=0A> =0A> =3D= =3D=3D=3D=3D The benefits of SQM =3D=3D=3D=3D=3D=3D=0A> =0A> Wikipedia sez= =E2=80=A6 SQM may refer to:=0A> =0A> - Sociedad Qu=C3=ADmica y Minera de Ch= ile - a Chilean mining and chemical=0A> enterprise=0A> - Software quality m= anagement=0A> - Spectrum quality management=0A> - Supplier Quality Manageme= nt=0A> - Sensors Quality Management Inc. - provides unbiased evaluations of= a company's=0A> operations relating to issues of quality, service, cleanli= ness and value=0A> - Sky Quality Meter, a device for measuring light pollut= ion=0A> =0A> and also : S=C3=A3o Miguel do Araguaia airport IATA code=0A> = =0A> sqm may refer to :=0A> =0A> - square metre=0A> - Windows Live Messenge= r log file extension=0A> =0A> So it doesn=E2=80=99t appear that there are a= ny seriously conflicting uses of that=0A> TLA=E2=80=A6=0A> =0A> And I prefe= r =E2=80=9CSmart Queue Management=E2=80=9D to =E2=80=9CSmarter Queue=0A> Ma= nagement=E2=80=9D. We=E2=80=99ll leave to someone else to go out on the wea= k branch=0A> and espouse =E2=80=9CSmarter queue management=E2=80=9D and =E2= =80=9CSmartest queue=0A> management=E2=80=9D. (What comes after that? =E2= =80=9CSmart and a half=E2=80=9D,=0A> Smart**2"?)=0A> =0A> Rich=0A> ________= _______________________________________=0A> Cerowrt-devel mailing list=0A> = Cerowrt-devel@lists.bufferbloat.net=0A> https://lists.bufferbloat.net/listi= nfo/cerowrt-devel=0A> ------=_20131220162230000000_26721 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable

Given that= there is no likelihood of making localized queue management "intelligent" = because it has no global information whatsoever, I strongly suggest that "s= mart" "intelligent" and even "active" are hugely misleading.

=0A

 

=0A

They= are based on a completely false premise - that queues should be allowed to= build at all, and that local information can solve highly transient global= problems.

=0A

 

=0A

"Dumb Queue Management" is going to be far superior. &= nbsp;Keep the queue at zero length, and try to be fair.

=0A

 

=0A

There's a= simple way to do the latter - use a filter (similar to a Bloom filter) tha= t captures recent/frequent users of the queue, and when the queue on an out= bound link grows more than about 2-3 packets (double buffering is all you n= eed to keep the link full) discard the most recent and frequent packets (or= send information that tells them to slow down).

=0A

 

=0A

There's been a l= ot of wasted time and effort trying to build queues long enough so that you= can be "intelligent", but by then you have already lost the battle.  = You've gotten into a positive feedback loop where you have encouraged the e= ndpoints to send more packets than you can ever drain out of the queue.

= =0A

 

=0A

I truly, truly do not understand why people don't look at realistic = network loads and structures.

=0A

 =

=0A=0A<= p style=3D"margin:0;padding:0;">

On Friday, December 20, 2013 3:= 52pm, "Rich Brown" <richb.hanover@gmail.com> said:

=0A=
=0A

> D= ave,
>
> You wrote:
>
> > What's in a = name? AQM has been pretty thoroughly defined to equal
> > active= queue *length* management and not packet scheduling.
> > Overlo= ading "AQM" what cerowrt does is apt to cause even more
> > conf= usion in the field than it already does. We discussed using LBO as
>= ; > a word but that appears hopelessly overloaded with leveraged buy out= .
> >
> > I go back to one I liked a while back:
> >
> > Smart Queue Management. (SQM)
> >
> > This got dissed on the aqm list too, but so far a viable alterna= tive
> > TLA has not appeared. It's sufficiently different to ha= ng a different
> > definition off of ("Smart queue management is= an intelligent
> > combination of better packet scheduling (flo= w queuing) techniques
> > along with with active queue length ma= nagement (aqm)=E2=80=9D)
>
> and
>
> >= Any ideas for a name for packet scheduling, prioritization, and active
> > queue management better than just "AQM", or "QoS"?
> &g= t;
> > SQM "Smarter Queue Management"
> > CeroShaper<= br />> > LBO Latency and Bandwidth Optimisation
>
> = I was prepared to agree with =E2=80=9CSQM=E2=80=9D, and had written a long = note
> (below) when my brain uttered =E2=80=9CIntelligent Queue Man= agement=E2=80=9D.
> I=E2=80=99m not convinced that one is better th= an the other=E2=80=A6
>
> Rich
>
> =3D=3D= =3D=3D=3D The benefits of SQM =3D=3D=3D=3D=3D=3D
>
> Wikip= edia sez=E2=80=A6 SQM may refer to:
>
> - Sociedad Qu=C3= =ADmica y Minera de Chile - a Chilean mining and chemical
> enterpr= ise
> - Software quality management
> - Spectrum quality ma= nagement
> - Supplier Quality Management
> - Sensors Qualit= y Management Inc. - provides unbiased evaluations of a company's
> = operations relating to issues of quality, service, cleanliness and value> - Sky Quality Meter, a device for measuring light pollution
&g= t;
> and also : S=C3=A3o Miguel do Araguaia airport IATA code
>
> sqm may refer to :
>
> - square metre
> - Windows Live Messenger log file extension
>
> So = it doesn=E2=80=99t appear that there are any seriously conflicting uses of = that
> TLA=E2=80=A6
>
> And I prefer =E2=80=9CSmar= t Queue Management=E2=80=9D to =E2=80=9CSmarter Queue
> Management= =E2=80=9D. We=E2=80=99ll leave to someone else to go out on the weak branch=
> and espouse =E2=80=9CSmarter queue management=E2=80=9D and =E2= =80=9CSmartest queue
> management=E2=80=9D. (What comes after that?= =E2=80=9CSmart and a half=E2=80=9D,
> Smart**2"?)
>
= > Rich
> _______________________________________________
&g= t; Cerowrt-devel mailing list
> Cerowrt-devel@lists.bufferbloat.net=
> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>=0A

 ------=_20131220162230000000_26721--