From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mout.gmx.net (mout.gmx.net [212.227.17.20]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id 655BB3BA8E for ; Sun, 3 Mar 2019 04:53:28 -0500 (EST) Received: from [192.168.1.133] ([77.182.35.26]) by mail.gmx.com (mrgmx102 [212.227.17.168]) with ESMTPSA (Nemesis) id 0MZOan-1ghU2y3fW1-00LBYb; Sun, 03 Mar 2019 10:53:26 +0100 Date: Sun, 03 Mar 2019 10:53:23 +0100 User-Agent: K-9 Mail for Android In-Reply-To: References: <72193310-7502-47B8-9554-7F8F9FA23204@heistp.net> <874l8mn9iy.fsf@toke.dk> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----ZAEQN1A3CS864Q2W9RAYYE5RGXDP7D" Content-Transfer-Encoding: 7bit To: cake@lists.bufferbloat.net, Pete Heist , George Amanakis CC: Cake List From: Sebastian Moeller Message-ID: <417E17B2-17F7-4106-A92D-C5B5AC41D808@gmx.de> X-Provags-ID: V03:K1:YADdA5inmh/Oz20VOFC9+YdxUWhlxy2mRrioc1HdMunDkkf7Kwv e4ct4lZVJisZB/J5GbCKOJkquJeIvHL0aC6cAlQcPCNmDgHWo0wKUikxrl2FsVj3xS+rw5F 8ky99+KiB4C9J7qf51F3KU+KvYmnz74lS0f6NvaO85AaJ51sPw0MYrLcAZmXt02DI/MY20F zlfhZj1zH5ig/02A8ngbg== X-Spam-Flag: NO X-UI-Out-Filterresults: notjunk:1;V03:K0:RmKYnxURCfs=:wH6bNTE6t7sWwJbH4uMKWb a6mbVUuPBDgwiO3L2SGonyTLtqHhSf+Q+8Ua4wYO3e/tejiDUl1bkeFDDFzozhAczhb2s6BX3 cCCUhY8g6hmfczXSQyIgH6vsSbTGU2g8AD0pfp21qb/syq/6kJqTfn0RczScFG7j+dfUo838x 4/Kv0UfvThSqguPf03EbI5WQBUuhCrJBJQd5HUaylpV+VS1giqYcGuxKhLh5lHmjiHC2MGE9q poOQppkBJfL5vb+cuff686MOXZnf+lh3xBituwLxo/qLkWu50rLqU0/1y3QmcxC87ADhdyAwV gKk7HDuIWKg6esVuLraiMUoPZ7TB/yG0cykXcivDtMY7q6yHxT7VcBfGTVCstERhIBagfDFoG SINhIQR6vYQEE6NvLiCx4wOM3eUQGcB0uE4UqHEDMS30U+UGMvkDT50FfZzF1/we6dgaLvp3w TqOq1FsDJE9yCrYriElknY2+zSKgd4zd/Zig0utVI30AlgGEyGbHe1LUSX0ugBzlF24azHxCW MYBsG1o8giznh5Jz71BQZGJX5gwH0X+nnSzXpbCwnLKGX+YgedTnebSxlS9i/Lho0UwjbSNkh GNnLhsDmQCF/0OKhtYtGVMGDEKOVMkr9VKYVXhRoxTHf4mFrYSS5m7+c1fMPeTf1hvtztd7Fl QTa4lT7ewcmcqumZo1UJRAKNmL8gW2lWOhfSGvwevgUiLQ+D/9M8dof5C3tzZiR18Utg3nELR w4RLiU072pTNDRMRU3di4l7WlU5+p6hDGjA47ZnZwcjD51LjQMnNG1ltmsqUbUY4I/uobVhzt RxyCAWlgoSybCCDkuAzBJyCzH+mHt9rNC3yXpisw020WLxUdRKPDSyhy7vdWGdZfVYysxgKMW aOe2nzet1EaW/10e2RmE8SDrbp5diwzFOzck6ym0ppdRkT1aPQg8HVarsuM+3MU5+qQFsutmm /LE1ne3+oDQ== Subject: Re: [Cake] Upstream submission of dual-mode fairness patch X-BeenThere: cake@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: Cake - FQ_codel the next generation List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 03 Mar 2019 09:53:28 -0000 ------ZAEQN1A3CS864Q2W9RAYYE5RGXDP7D Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Mmmh, Doesn't this look like ingress magic being applied selectively to the user= s based on number of flows? I thought that the idea behind the ingress keyw= ord is to effectively shape harder the more bulk flows are coming in=2E=20 Just asking=2E=2E=2E Best regards Sebastian On March 3, 2019 8:19:00 AM GMT+01:00, Pete Heist wrot= e: >Oops, sorry I didn=E2=80=99t notice this before but it=E2=80=99s the ingr= ess keyword >that makes the difference: > >qdisc cake 802c: dev ifbLink2 parent 1:1 bandwidth unlimited besteffort >dual-dsthost nonat nowash ingress no-ack-filter no-split-gso rtt >100=2E0ms raw overhead 0=20 > >IP1, 1 up: 46=2E8 Mbit >IP2, 16 up: 46=2E8 Mbit >IP1, 16 down: 41=2E9 Mbit >IP2, 1 down: 51=2E1 Mbit > >qdisc cake 803a: dev ifbLink2 parent 1:1 bandwidth unlimited besteffort >dual-dsthost nonat nowash no-ack-filter no-split-gso rtt 100=2E0ms raw >overhead 0=20 > >IP1, 1 up: 46=2E6 Mbit >IP2, 16 up: 46=2E5 Mbit >IP1, 16 down: 46=2E4 Mbit >IP2, 1 down: 46=2E5 Mbit > >Also, my setup is different in that I use cake as a leaf to either hfsc >or htb=2E But, it=E2=80=99s adding the ingress keyword that causes the im= balance >for me=2E > >> On Mar 2, 2019, at 11:20 AM, Pete Heist wrote: >>=20 >> Great, thanks for trying it=2E That strongly suggests a problem with >the kernel (or driver) I=E2=80=99m using=2E Feels better to know it works= as it >should in recent kernels though=2E=2E=2E >>=20 >>> On Mar 2, 2019, at 5:47 AM, George Amanakis >wrote: >>>=20 >>> On Fri, 2019-03-01 at 22:02 -0500, George Amanakis wrote: >>>>=20 >>>> I will setup a vlan and try again=2E >>>>=20 >>>=20 >>> I replicated Pete's VLAN setup, and I am getting fairness: >>>=20 >>> IP1,2 <---> router enp1s0 / router enp1s0=2E100 <---> server >>>=20 >>> IP1, 1 up: 46=2E73 mbit/s >>> IP2, 32 up: 46=2E91 >>> IP1, 32 down: 46=2E70 >>> IP2, 1 down: 46=2E89 >>>=20 >>> qdisc cake 8006: dev enp1s0=2E100 root refcnt 2 bandwidth 100Mbit >>> diffserv3 dual-srchost nonat nowash no-ack-filter split-gso nofwmark >>> rtt 100=2E0ms noatm overhead 4 >>>=20 >>> qdisc cake 8002: dev ifb0 root refcnt 2 bandwidth 100Mbit diffserv3 >>> dual-dsthost nonat nowash no-ack-filter split-gso nofwmark rtt >100=2E0ms >>> noatm overhead 4 >>>=20 >>>=20 >>>=20 >>=20 --=20 Sent from my Android device with K-9 Mail=2E Please excuse my brevity=2E ------ZAEQN1A3CS864Q2W9RAYYE5RGXDP7D Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable Mmmh,

Doesn't this = look like ingress magic being applied selectively to the users based on num= ber of flows? I thought that the idea behind the ingress keyword is to effe= ctively shape harder the more bulk flows are coming in=2E

Just aski= ng=2E=2E=2E

Best regards
Sebastian

On March 3, 2019 8:19:00 AM GMT+01:00, Pete Heist <pete@heistp=2Enet= > wrote:
Oops, sorry I didn=E2=80=99t notice this before but it=E2=80=99s the ingre= ss keyword that makes the difference:

<= div class=3D"">qdisc cake 802c: dev ifbLink2 parent 1:1 bandwidth unlimited= besteffort dual-dsthost nonat nowash ingress no-ack-filter no-split-gso rt= t 100=2E0ms raw overhead 0 

=
IP1, 1 up: 46=2E8 Mbit
IP2, 16 up: 46= =2E8 Mbit
IP1, 16 down: 41=2E9 Mbit
IP2, 1 down: 51=2E1 Mbit

<= /div>
qdisc ca= ke 803a: dev ifbLink2 parent 1:1 bandwidth unlimited besteffort dual-dsthos= t nonat nowash no-ack-filter no-split-gso rtt 100=2E0ms raw overhead 0 = ;

IP1, 1 u= p: 46=2E6 Mbit
IP2, 16 up: 46=2E5 Mbit
<= div>
IP1, 16 down: 46=2E4 Mbit
IP2, 1 d= own: 46=2E5 Mbit

= Also, my setup is different in that I use cake as a leaf to either hfsc or = htb=2E But, it=E2=80=99s adding the ingress keyword that causes the imbalan= ce for me=2E

On Mar 2, 2019, at 11:20= AM, Pete Heist <pete@he= istp=2Enet> wrote:

Great, thanks for trying it=2E That strongly su= ggests a problem with the kernel (or driver) I=E2=80=99m using=2E Feels bet= ter to know it works as it should in recent kernels though=2E=2E=2E

On Mar 2, 2019, = at 5:47 AM, George Amanakis <gamanakis@gmail=2Ecom> wrote:

= On Fri, 2019-03-01 at 22:02 -0500, George Amanakis wrote:

I will setup a vlan and tr= y again=2E


I repli= cated Pete's VLAN setup, and I am getting fairness:

IP1,2 <---> router enp1s0 / router enp1s0=2E100 <---> ser= ver

IP1, 1 up:      4= 6=2E73 mbit/s
IP2, 32 up:     46=2E91
IP1, 32 down:   46=2E70
IP2, 1 down: &nb= sp;  46=2E89

qdisc cake 8006: dev en= p1s0=2E100 root refcnt 2 bandwidth 100Mbit
diffserv3 dual-src= host nonat nowash no-ack-filter split-gso nofwmark
rtt 100=2E= 0ms noatm overhead 4

qdisc cake 8002: dev ifb0= root refcnt 2 bandwidth 100Mbit diffserv3
dual-dsthost nonat= nowash no-ack-filter split-gso nofwmark rtt 100=2E0ms
noatm = overhead 4





=
--
Sent from my Android device with K-9 Mail=2E = Please excuse my brevity=2E ------ZAEQN1A3CS864Q2W9RAYYE5RGXDP7D--