From: "Toke Høiland-Jørgensen" <toke@toke.dk>
To: Mikael Abrahamsson <swmike@swm.pp.se>
Cc: "bloat@lists.bufferbloat.net" <bloat@lists.bufferbloat.net>
Subject: Re: [Bloat] RED against bufferbloat
Date: Wed, 25 Feb 2015 12:04:35 +0100 [thread overview]
Message-ID: <87pp8yfe0s.fsf@toke.dk> (raw)
In-Reply-To: <alpine.DEB.2.02.1502251138490.4007@uplift.swm.pp.se> (Mikael Abrahamsson's message of "Wed, 25 Feb 2015 11:47:53 +0100 (CET)")
Mikael Abrahamsson <swmike@swm.pp.se> writes:
> To me, FQ is important especially for lower speeds. This also maps
> well into the CPU based nature of FQ (that it's hard to do in
> "silicon").
Well my research has been mainly focused on the consumer / edge case.
I.e. things a Linux box can drive in hardware, so up to 100Mbit to a
Gbit depending on the size of the box. In these cases, fq_codel adds no
measurable overhead compared to the CPU load of just pushing the
packets.
> We're not going to see FQ_CODEL on a 200 interface large router
> designed to push 100 gigabit/s of traffic, at least not in any
> interesting price point.
Maybe not; but note that recent work in the Linux kernel makes it quite
capable of pushing 40GigE, and I believe sched_fq (which does 'real'
fairness queueing with a queue per flow, not hash buckets as fq_codel
does) has been shown to scale to millions of concurrent flows.
> Can we do bidirectional traffic FQ_CODEL on the CPE to try to achieve
> basically zero loss in the AR policer for any normal kind of use
> scenario using TCP. I have not seen any simulations looking at this.
Well this is basically what we're doing with the SQM scripts in CeroWRT:
put a software rate limiter in the CPE and configure it to a slightly
lower value than whatever the cable/DSL modem runs at. This works quite
well, but the software rate limiter is fairly CPU hungry, so small boxes
struggle. I had to substitute a full x86 box for my Netgear router to
run at 100Mbit for my home connection.
> Because my belief is that even though we might say we love FQ here,
> we're not going to see high speed FQ on higher end ISP aggregation
> routers. So if we want FQ, we're going to have to do it in the CPEs.
Well OpenWRT turns on fq_codel by default now. :)
And yeah, I'm not holding by breath for big iron vendors to include
fq_codel in their products. But that doesn't mean it can't go into CPEs.
And end hosts and access points. And inside tunnels (VPN daemons for
instance).
-Toke
next prev parent reply other threads:[~2015-02-25 11:04 UTC|newest]
Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-24 15:43 sahil grover
2015-02-24 16:13 ` Matt Mathis
2015-02-24 22:39 ` Kathleen Nichols
2015-02-25 6:46 ` Mikael Abrahamsson
2015-02-25 6:54 ` David Lang
2015-02-25 6:59 ` Mikael Abrahamsson
2015-02-25 8:29 ` Alex Elsayed
2015-02-25 8:06 ` Bob Briscoe
2015-02-25 8:42 ` Alex Elsayed
2015-02-25 9:18 ` Michael Welzl
2015-02-25 9:29 ` Sebastian Moeller
2015-02-25 10:10 ` Michael Welzl
2015-02-25 10:24 ` Toke Høiland-Jørgensen
2015-02-25 10:47 ` Mikael Abrahamsson
2015-02-25 11:04 ` Toke Høiland-Jørgensen [this message]
2015-02-25 18:39 ` Bill Ver Steeg (versteb)
2015-02-26 9:01 ` MUSCARIELLO Luca IMT/OLN
2015-02-26 10:39 ` Mikael Abrahamsson
2015-02-26 10:41 ` Toke Høiland-Jørgensen
2015-02-26 10:44 ` Mikael Abrahamsson
2015-02-26 10:51 ` Toke Høiland-Jørgensen
2015-02-26 10:59 ` Sebastian Moeller
2015-02-26 11:12 ` Jonathan Morton
2015-02-27 0:26 ` Dave Taht
2015-02-26 10:45 ` Sebastian Moeller
2015-02-26 11:34 ` Jonathan Morton
2015-02-26 12:59 ` Mikael Abrahamsson
2015-02-26 11:26 ` MUSCARIELLO Luca IMT/OLN
2015-02-26 12:57 ` Mikael Abrahamsson
2015-02-25 13:25 ` Sebastian Moeller
2015-02-25 13:36 ` Mikael Abrahamsson
2015-02-25 13:38 ` Toke Høiland-Jørgensen
2015-02-25 14:05 ` Mikael Abrahamsson
2015-02-25 18:51 ` Bill Ver Steeg (versteb)
2015-02-25 14:16 ` MUSCARIELLO Luca IMT/OLN
2015-02-25 16:09 ` Mikael Abrahamsson
2015-02-25 17:34 ` MUSCARIELLO Luca IMT/OLN
2015-02-25 17:56 ` Jonathan Morton
2015-02-26 12:54 ` Mikael Abrahamsson
2015-02-26 14:06 ` MUSCARIELLO Luca IMT/OLN
2015-02-26 14:18 ` Mikael Abrahamsson
2015-02-26 15:18 ` MUSCARIELLO Luca IMT/OLN
2015-02-26 17:04 ` Dave Taht
2015-02-26 18:07 ` Dave Taht
2015-02-26 18:33 ` [Bloat] RE : " luca.muscariello
2015-02-26 18:59 ` [Bloat] " Mikael Abrahamsson
2015-02-26 19:44 ` Bill Ver Steeg (versteb)
2015-02-26 20:42 ` Jonathan Morton
2015-02-26 21:50 ` Dave Taht
2015-02-25 16:54 ` Sebastian Moeller
2015-02-25 10:54 ` Michael Welzl
2015-02-25 11:24 ` Toke Høiland-Jørgensen
2015-02-25 12:08 ` Jonathan Morton
2015-02-25 19:04 ` David Lang
2015-02-25 19:30 ` Michael Welzl
2015-02-25 9:31 ` Alex Elsayed
2015-02-25 10:37 ` Michael Welzl
2015-02-25 10:54 ` Alex Elsayed
2015-02-25 17:28 ` Bob Briscoe
2015-02-25 18:03 ` Dave Taht
2015-02-26 9:36 ` Sebastian Moeller
2015-02-25 17:57 ` Dave Taht
2015-02-25 19:25 Hal Murray
2015-02-25 20:00 ` Jonathan Morton
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: https://lists.bufferbloat.net/postorius/lists/bloat.lists.bufferbloat.net/
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87pp8yfe0s.fsf@toke.dk \
--to=toke@toke.dk \
--cc=bloat@lists.bufferbloat.net \
--cc=swmike@swm.pp.se \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox