From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-s66.mailgun.info (mail-s66.mailgun.info [184.173.153.194]) (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 BFE303B2A4 for ; Thu, 28 Mar 2019 16:45:04 -0400 (EDT) DKIM-Signature: a=rsa-sha256; v=1; c=relaxed/relaxed; d=flamingpc.com; q=dns/txt; s=pic; t=1553805904; h=Content-Type: Cc: To: Subject: Message-ID: Date: From: In-Reply-To: References: MIME-Version: Sender; bh=5txNgVrzzkr+Q8EC+vs1JsCDqlVuS08ilzynKqJDlV8=; b=VBVtm95o7uPNnfphsYBfIC8ANrfEyvaOishhZvGKZ9asJ1vzP7ZyQRVBRGl7XF/fZ4TEOoeg 4BBdPjA6mJ0cFW1P11wBPcnjjuhoji/2Lb6ijiPHShNHkQPmIdY43uAIYVPyL1vvJNK4mr+g NyJsaQTGqkU9DSoobBuF/PjIlo4= X-Mailgun-Sending-Ip: 184.173.153.194 X-Mailgun-Sid: WyJiMjA5OSIsICJibG9hdEBsaXN0cy5idWZmZXJibG9hdC5uZXQiLCAiNjY0ODllNyJd Sender: bufferbloat@flamingpc.com Received: from mail-vs1-f46.google.com (mail-vs1-f46.google.com [209.85.217.46]) by mxa.mailgun.org with ESMTP id 5c9d324f.7f55586d8fb0-smtp-out-n01; Thu, 28 Mar 2019 20:45:03 -0000 (UTC) Received: by mail-vs1-f46.google.com with SMTP id n4so34161vsm.3 for ; Thu, 28 Mar 2019 13:45:03 -0700 (PDT) X-Gm-Message-State: APjAAAUlnOjcpbACGiu0gC0pmGDM5J6lD/2H8HKBXXWGw/yAANY834KA xaAXKw1l+TQS5nZckP2pJPdGV63yr9qPeniNbqc= X-Google-Smtp-Source: APXvYqwv25V1Sdn/ZhMi+z7j+mwFZO+QDih959o3RTSR8yfdMl0ufMOFouMSPCzJVIuvNW5g86LilN7NPNwjwN8NV5E= X-Received: by 2002:a67:e881:: with SMTP id x1mr26199093vsn.48.1553805903056; Thu, 28 Mar 2019 13:45:03 -0700 (PDT) MIME-Version: 1.0 References: <2F486932-0989-498D-A2CD-2E6EB520A758@gmx.de> In-Reply-To: <2F486932-0989-498D-A2CD-2E6EB520A758@gmx.de> From: cloneman Date: Thu, 28 Mar 2019 16:44:52 -0400 X-Gmail-Original-Message-ID: Message-ID: To: Sebastian Moeller Cc: bloat , dave.taht@gmail.com Content-Type: multipart/alternative; boundary="000000000000e6d6fb05852da0a4" Subject: Re: [Bloat] unfixable bloat on Cable modem CODA-4680 (puma 7) X-BeenThere: bloat@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: General list for discussing Bufferbloat List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 28 Mar 2019 20:45:04 -0000 --000000000000e6d6fb05852da0a4 Content-Type: text/plain; charset="UTF-8" I have discovered the source of the confusion. This is different, but similar to PUMA6 problems, I think. Certain types of traffic are getting de-prioritized, especially small packets under load. Comparison: Normal ping: Minimum = 17ms, Maximum = 734ms, *Average = 202ms* Large 1472-byte ping Minimum = 18ms, Maximum = 229ms, *Average = 47ms* TCP-Based ping via psping.exe Minimum = 14.15ms, Maximum = 40.39ms, *Average = 25.88ms* For more fun: ICMP in a VPN, ping set to 13-bytes Minimum = 36ms, Maximum = 602ms, *Average = 208ms* ICMP in a VPN, 1300 bytes Minimum = 24ms, Maximum = 72ms, *Average = 40ms* This explains why certain synthetic bufferbloat benchmarks were passing with flying colors but my basic ICMP ping was terrible under load. VPN does not seem to help, so it's not packet inspection... it's failure of tiny packets and success of larger, normal ones. The problem only occurs during download stress, not upload stress. Anyway, I guess this is mostly an FYI. Perhaps a script that does variable packet sizes for the generated small flow traffic could be useful to someone in the future. On Thu, Mar 28, 2019 at 8:03 AM Sebastian Moeller wrote: > HI cloneman, > > maybe have a look at http://www.dslreports.com/tools/puma6? > > > On Mar 28, 2019, at 12:00, cloneman wrote: > > > > I'm getting a weird problem with this new modem , it's Docsis 3.1 > > > > latency/jitter measured via ping is about ~300ms during download > pressure of 3 http transfers. The latency doesn't appear on synthetic > bufferbloat benchmarks, though. (dslreports ; 20ms ("A") > > > > question: trying to figure out why I'm seeing this particular behavior, > and determine if it'd a real issue or just an anomalie with ICMP traffic > > > > I've not seen ping do this with other modems. > > > > Perhaps I need to try some other traffic generation/ capture tools? > > > > > > _______________________________________________ > > Bloat mailing list > > Bloat@lists.bufferbloat.net > > https://lists.bufferbloat.net/listinfo/bloat > --000000000000e6d6fb05852da0a4 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
I have discovered the source of= the confusion. This is different, but similar to=C2=A0 PUMA6 problems, I t= hink.
Certain types of traffic are getting de-pr= ioritized, especially small packets under load.

Comparison:

Normal ping:=C2=A0
=C2=A0M= inimum =3D 17ms, Maximum =3D 734ms, Average =3D 202ms
=
Large 1472-byte ping
=C2=A0Minimum =3D 18ms= , Maximum =3D 229ms, Average =3D 47ms

T= CP-Based ping via psping.exe
=C2=A0Minimum =3D 14.15ms, Maximum = =3D 40.39ms, Average =3D 25.88ms

For more fun:

ICMP in a VPN, ping set to 1= 3-bytes
Minimum =3D 36ms, Maximum =3D 602ms, Average =3D 208ms=

ICMP in a VPN, 1300 bytes
=C2= =A0Minimum =3D 24ms, Maximum =3D 72ms, Average =3D 40ms


This explains why certain synthetic bufferb= loat benchmarks were passing with flying colors but my basic ICMP ping was = terrible under load. VPN does not seem to help, so it's not packet insp= ection... it's failure of tiny packets and success of larger, normal on= es.


The problem only occurs during = download stress, not upload stress. Anyway, I guess this is mostly an FYI.<= br>Perhaps=C2=A0 a script that=C2=A0 does variable packet sizes for=C2=A0th= e generated small flow traffic could be useful to someone in the future.




On Thu, Mar 28, 2019 at 8:03 AM Sebastian Moeller <moeller0@gmx.de> wrote:
HI cloneman,

maybe have a look at http://www.dslreports.com/tools/puma6= ?

> On Mar 28, 2019, at 12:00, cloneman <bufferbloat@flamingpc.com> wrote: >
> I'm getting a weird problem with this new modem , it's Docsis = 3.1
>
> latency/jitter measured via ping is about ~300ms during download press= ure of 3 http transfers. The latency doesn't appear on synthetic buffer= bloat benchmarks, though. (dslreports ; 20ms ("A")
>
> question: trying to figure out why I'm seeing this particular beha= vior, and determine if it'd a real issue or just an anomalie with ICMP = traffic
>
> I've not seen ping do this with other modems.
>
> Perhaps I need to try some other traffic generation/ capture=C2=A0 too= ls?
>
>
> _______________________________________________
> Bloat mailing list
> Bloat= @lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
--000000000000e6d6fb05852da0a4--