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 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by huchra.bufferbloat.net (Postfix) with ESMTPS id A705421F71F for ; Wed, 1 Jul 2015 08:37:39 -0700 (PDT) Received: from smtp13.relay.iad3a.emailsrvr.com (localhost.localdomain [127.0.0.1]) by smtp13.relay.iad3a.emailsrvr.com (SMTP Server) with ESMTP id 20000100A23; Wed, 1 Jul 2015 11:37:38 -0400 (EDT) Received: from app30.wa-webapps.iad3a (relay-webapps.rsapps.net [172.27.255.140]) by smtp13.relay.iad3a.emailsrvr.com (SMTP Server) with ESMTP id 03E771009F1; Wed, 1 Jul 2015 11:37:37 -0400 (EDT) X-Sender-Id: dpreed@reed.com Received: from app30.wa-webapps.iad3a (relay-webapps.rsapps.net [172.27.255.140]) by 0.0.0.0:25 (trex/5.4.2); Wed, 01 Jul 2015 15:37:38 GMT Received: from reed.com (localhost.localdomain [127.0.0.1]) by app30.wa-webapps.iad3a (Postfix) with ESMTP id E53708003E; Wed, 1 Jul 2015 11:37:37 -0400 (EDT) Received: by apps.rackspace.com (Authenticated sender: dpreed@reed.com, from: dpreed@reed.com) with HTTP; Wed, 1 Jul 2015 11:37:37 -0400 (EDT) Date: Wed, 1 Jul 2015 11:37:37 -0400 (EDT) From: dpreed@reed.com To: "Mikael Abrahamsson" MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_20150701113737000000_83375" Importance: Normal X-Priority: 3 (Normal) X-Type: html In-Reply-To: References: <8B853F1C-DE5D-4F3D-88CC-CB8DA2D3E8B1@gmx.de> <04331509-F163-4184-90B4-8589073AFD62@gmx.de> <09BA156C-460D-4794-A082-33E805F3D6FD@gmx.de> <5436B48C-0803-46DA-B355-14E917A5BB37@gmx.de> <4E002218-174D-44F9-91A0-C7F34B9E83C7@gmx.de> <87pp4eomfx.fsf@alrua-karlstad.karlstad.toke.dk> <1435585587.97486240@apps.rackspace.com> <1435681242.435511455@apps.rackspace.com> X-Auth-ID: dpreed@reed.com Message-ID: <1435765057.937422821@apps.rackspace.com> X-Mailer: webmail/11.5.2-RC Cc: "cerowrt-devel@lists.bufferbloat.net" Subject: Re: [Cerowrt-devel] Build instructions for regular OpenWRT with Ceropackages 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: Wed, 01 Jul 2015 15:38:08 -0000 ------=_20150701113737000000_83375 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable =0AMikael, very very helpful, thanks.=0A =0AI now understand what you are t= rying to prove/test in your experiments, but there is definitely a need for= cake when the dominant use is hi-bitrate WiFi (AC1900) talking to one or m= ore 1 GigE wired paths. And hi bitrate WiFi itself has significantly varia= ble rate capability so it probably needs more feedback than cake might prov= ide to deal with variability.=0A =0ASince 100+ Mb/sec is supplied by many I= nternet Access Providers now, it's timely to be able to process packets co= ming at those rates on the wireline side carried over 1 GigE (my provider, = RCN, claims to offer 110+ here in Needham, but with an odd requirement that= I buy their router if I get that service - I am trying to get to the botto= m of what that is before I upgrade or switch to one of the two other provid= ers, Comcast and Verizon. Maybe it is just that they want customers to not = get screwed up if they have a router without 1GigE WAN adapter, complaining= that they can't get 110.).=0A =0AI'd like to see both whatever I get from = the IAP, and also what my in-home NAS can provide, along with other service= s.=0A=0A=0AOn Tuesday, June 30, 2015 3:58pm, "Mikael Abrahamsson" said:=0A=0A=0A=0A> On Tue, 30 Jun 2015, dpreed@reed.com wrote:=0A= > =0A> > What happens if the SoC ports aren't saturated, but the link is Gi= gE?=0A> > That is, suppose this is an access link to a GigE home or office = LAN=0A> > with wired servers?=0A> =0A> As far as I can tell, the device loo= ks like this:=0A> =0A> wifi2------=0A> wifi1----\|=0A> SOC2 6-|=0A> SOC1 5-= |=0A> WAN 4-|=0A> LAN1 3-| (switch)=0A> LAN2 2-|=0A> LAN3 1-|=0A> LAN4 0-|= =0A> =0A> LAN1-4 and SOC2 is in one vlan, and SOC1 and WAN is in a second v= lan. This=0A> basically means there is no way to get traffic into SOC1 that= goes out=0A> SOC2 that will saturate either port, because they're both gig= e. Only way=0A> to saturate the SOC port would be if the SOC itself "create= d" traffic, for=0A> instance by being a fileserver, or if there is signific= ant traffic on the=0A> wifi (which has PCI-E connectivity).=0A> =0A> So it'= s impossible to congest SOC1 or SOC2 (egress) by running traffic=0A> LAN<->= WAN alone.=0A> =0A> --=0A> Mikael Abrahamsson email: swmike@swm.pp.se=0A> ------=_20150701113737000000_83375 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable

Mika= el, very very helpful, thanks.

=0A

 =0A

I now understand what you are trying to pr= ove/test in your experiments, but there is definitely a need for cake = when the dominant use is hi-bitrate WiFi (AC1900) talking to one or more 1 = GigE wired paths.  And hi bitrate WiFi itself has significantly variab= le rate capability so it probably needs more feedback than cake might = provide to deal with variability.

=0A

 =

=0A

Since 100+ Mb/sec is supplied by many I= nternet Access Providers now,  it's timely to be able to process packe= ts coming at those rates on the wireline side carried over 1 GigE (my provi= der, RCN, claims to offer 110+ here in Needham, but with an odd requirement= that I buy their router if I get that service - I am trying to get to the = bottom of what that is before I upgrade or switch to one of the two other p= roviders, Comcast and Verizon. Maybe it is just that they want customers to= not get screwed up if they have a router without 1GigE WAN adapter, compla= ining that they can't get 110.).

=0A

 <= /p>=0A

I'd like to see both whatever I get from= the IAP, and also what my in-home NAS can provide, along with other servic= es.

=0A= =0A



On Tuesday, June 30, 2015 3:58p= m, "Mikael Abrahamsson" <swmike@swm.pp.se> said:

=0A=0A

> On Tue,= 30 Jun 2015, dpreed@reed.com wrote:
>
> > What happens= if the SoC ports aren't saturated, but the link is GigE?
> > Th= at is, suppose this is an access link to a GigE home or office LAN
>= ; > with wired servers?
>
> As far as I can tell, the d= evice looks like this:
>
> wifi2------
> wifi1----= \|
> SOC2 6-|
> SOC1 5-|
> WAN 4-|
> LAN1 3= -| (switch)
> LAN2 2-|
> LAN3 1-|
> LAN4 0-|
&= gt;
> LAN1-4 and SOC2 is in one vlan, and SOC1 and WAN is in a sec= ond vlan. This
> basically means there is no way to get traffic int= o SOC1 that goes out
> SOC2 that will saturate either port, because= they're both gige. Only way
> to saturate the SOC port would be if= the SOC itself "created" traffic, for
> instance by being a filese= rver, or if there is significant traffic on the
> wifi (which has P= CI-E connectivity).
>
> So it's impossible to congest SOC1= or SOC2 (egress) by running traffic
> LAN<->WAN alone.
= >
> --
> Mikael Abrahamsson email: swmike@swm.pp.se
>

=0A ------=_20150701113737000000_83375--