From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtp97.iad3a.emailsrvr.com (smtp97.iad3a.emailsrvr.com [173.203.187.97]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 8D46A3CB35 for ; Mon, 4 Dec 2017 14:59:57 -0500 (EST) Received: from smtp21.relay.iad3a.emailsrvr.com (localhost [127.0.0.1]) by smtp21.relay.iad3a.emailsrvr.com (SMTP Server) with ESMTP id 4A64D25CE7; Mon, 4 Dec 2017 14:59:57 -0500 (EST) Received: from app44.wa-webapps.iad3a (relay-webapps.rsapps.net [172.27.255.140]) by smtp21.relay.iad3a.emailsrvr.com (SMTP Server) with ESMTP id 26C0525CBC; Mon, 4 Dec 2017 14:59:57 -0500 (EST) X-Sender-Id: dpreed@reed.com Received: from app44.wa-webapps.iad3a (relay-webapps.rsapps.net [172.27.255.140]) by 0.0.0.0:25 (trex/5.7.12); Mon, 04 Dec 2017 14:59:57 -0500 Received: from reed.com (localhost.localdomain [127.0.0.1]) by app44.wa-webapps.iad3a (Postfix) with ESMTP id 16EF120069; Mon, 4 Dec 2017 14:59:57 -0500 (EST) Received: by apps.rackspace.com (Authenticated sender: dpreed@reed.com, from: dpreed@reed.com) with HTTP; Mon, 4 Dec 2017 14:59:57 -0500 (EST) X-Auth-ID: dpreed@reed.com Date: Mon, 4 Dec 2017 14:59:57 -0500 (EST) From: dpreed@reed.com To: "Mikael Abrahamsson" Cc: "Pedro Tumusok" , "cerowrt-devel@lists.bufferbloat.net" , "bloat" MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_20171204145957000000_38600" Importance: Normal X-Priority: 3 (Normal) X-Type: html In-Reply-To: References: <92906bd8-7bad-945d-83c8-a2f9598aac2c@lackof.org> <87bmjff7l6.fsf_-_@nemesis.taht.net> Message-ID: <1512417597.091724124@apps.rackspace.com> X-Mailer: webmail/12.9.10-RC Subject: Re: [Cerowrt-devel] [Bloat] DC behaviors today X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.20 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: Mon, 04 Dec 2017 19:59:57 -0000 ------=_20171204145957000000_38600 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable =0AI suggest we stop talking about throughput, which has been the mistaken = idea about networking for 30-40 years.=0A =0AAlmost all networking ends up = being about end-to-end response time in a multiplexed system.=0A =0AOr put = another way: "It's the Latency, Stupid".=0A =0AI get (and have come to expe= ct) 27 msec. RTT's under significant load, from Boston suburb to Sunnyvale,= CA.=0A =0AI get 2 microsecond RTT's within my house (using 10 GigE).=0A = =0AWhat will we expect tomorrow?=0A =0AThis is related to Bufferbloat, beca= use queueing delay is just not a good thing in these contexts - contexts wh= ere Latency Matters. We provision multiplexed networks based on "peak capac= ity" never being reached.=0A =0AConsequently, 1 Gig to the home is "table s= takes". And in DOCSIS 3.1 deployments that is what is being delivered, chea= p, today.=0A =0AAnd 10 Gig within the home is becoming "table stakes", espe= cially for applications that need quick response to human interaction.=0A = =0A1 NvME drive already delivers around 11 Gb/sec at its interface. That's = what is needed in the network to "impedance match".=0A =0A802.11ax already = gives around 10 Gb/sec. wireless (and will be on the market soon).=0A =0ATh= e folks who think that having 1 Gb/sec to the home would only be important = if you had to transfer at that rate 8 hours a day are just not thinking cle= arly about what "responsiveness" means.=0A =0AFor a different angle on this= , think about what the desirable "channel change time" is if a company like= Netflix were covering all the football (I mean US's soccer) games in the w= orld. You'd like to fill the "buffer" in 100 msec. so channel change to som= e new channel is responsive. 100 msec. of 4K sports, which you are watching= in "real time" needs to be buffered, and you want no more than a second or= two of delay from camera to your screen. So buffering up 1 second of a new= ly selected 4 K video stream in 100 msec. on demand is why you need such sp= eeds. Do the math.=0A =0AVR sports coverage - even moreso.=0A =0A =0A=0A=0A= On Monday, December 4, 2017 7:44am, "Mikael Abrahamsson" = said:=0A=0A=0A=0A> On Mon, 4 Dec 2017, Pedro Tumusok wrote:=0A> =0A> > Loo= king at chipsets coming/just arrived from the chipset vendors, I think=0A> = > we will see CPE with 10G SFP+ and 802.11ax Q3/Q4 this year.=0A> > Price i= s of course a bit steeper than the 15USD USB DSL modem :P, but=0A> > probab= ly fits nicely for the SMB segment.=0A> =0A> https://kb.netgear.com/31408/W= hat-SFP-modules-are-compatible-with-my-Nighthawk-X10-R9000-router=0A> =0A> = This has been available for a while now. Only use-case I see for it is=0A> = Comcast 2 gigabit/s service, that's the only one I know of that would fit= =0A> this product (since it has no downlink 10GE ports).=0A> =0A> --=0A> Mi= kael Abrahamsson email: swmike@swm.pp.se=0A> ______________________________= _________________=0A> Cerowrt-devel mailing list=0A> Cerowrt-devel@lists.bu= fferbloat.net=0A> https://lists.bufferbloat.net/listinfo/cerowrt-devel=0A> ------=_20171204145957000000_38600 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable

I suggest we stop talk= ing about throughput, which has been the mistaken idea about networking for= 30-40 years.

=0A

 

=0A

Almo= st all networking ends up being about end-to-end response time in a multipl= exed system.

=0A

 

=0A

Or pu= t another way: "It's the Latency, Stupid".

=0A

 <= /p>=0A

I get (and have come to expect) 27 msec. RTT's und= er significant load, from Boston suburb to Sunnyvale, CA.

=0A

 

=0A

I get 2 microsecond RTT's within my= house (using 10 GigE).

=0A

 

=0A

What will we expect tomorrow?

=0A

 

=0A<= p style=3D"margin:0;padding:0;font-family: verdana; font-size: 10pt; word-w= rap: break-word;">This is related to Bufferbloat, because queueing delay is= just not a good thing in these contexts - contexts where Latency Matters. = We provision multiplexed networks based on "peak capacity" never being reac= hed.

=0A

 

=0A

Consequently,= 1 Gig to the home is "table stakes". And in DOCSIS 3.1 deployments that is= what is being delivered, cheap, today.

=0A

 

= =0A

And 10 Gig within the home is becoming "table stakes"= , especially for applications that need quick response to human interaction= .

=0A

 

=0A

1 NvME drive alr= eady delivers around 11 Gb/sec at its interface. That's what is needed in t= he network to "impedance match".

=0A

 

=0A

802.11ax already gives around 10 Gb/sec. wireless (and will = be on the market soon).

=0A

 

=0A

The folks who think that having 1 Gb/sec to the home would only be im= portant if you had to transfer at that rate 8 hours a day are just not thin= king clearly about what "responsiveness" means.

=0A

&n= bsp;

=0A

For a different angle on this, think about wh= at the desirable "channel change time" is if a company like Netflix were co= vering all the football (I mean US's soccer) games in the world. You'd like= to fill the "buffer" in 100 msec. so channel change to some new channel is= responsive. 100 msec. of 4K sports, which you are watching in "real time" = needs to be buffered, and you want no more than a second or two of delay fr= om camera to your screen. So buffering up 1 second of a newly selected 4 K = video stream in 100 msec. on demand is why you need such speeds. Do the mat= h.

=0A

 

=0A

VR sports cover= age - even moreso.

=0A

 

=0A

 

=0A=0A



On Monday, December 4, 2017 7:44am, "M= ikael Abrahamsson" <swmike@swm.pp.se> said:

=0A
=0A

> On Mon, 4 Dec 2017, Pe= dro Tumusok wrote:
>
> > Looking at chipsets coming/jus= t arrived from the chipset vendors, I think
> > we will see CPE = with 10G SFP+ and 802.11ax Q3/Q4 this year.
> > Price is of cour= se a bit steeper than the 15USD USB DSL modem :P, but
> > probab= ly fits nicely for the SMB segment.
>
> https://kb.netgear= .com/31408/What-SFP-modules-are-compatible-with-my-Nighthawk-X10-R9000-rout= er
>
> This has been available for a while now. Only use-c= ase I see for it is
> Comcast 2 gigabit/s service, that's the only = one I know of that would fit
> this product (since it has no downli= nk 10GE ports).
>
> --
> Mikael Abrahamsson email:= swmike@swm.pp.se
> _______________________________________________=
> Cerowrt-devel mailing list
> Cerowrt-devel@lists.bufferb= loat.net
> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>

=0A
 ------=_20171204145957000000_38600--