[Bloat] buffer sizing meeting notes

[Bloat] buffer sizing meeting notes

[Dave Taht]

There's some good preso from various representatives in the dc market here.

The p4 stuff, in particular (from barefoot) is looking impressive.

http://yuba.stanford.edu/~bspang/buffer-sizing-meeting/


-- 

Dave Täht
CTO, TekLibre, LLC
http://www.teklibre.com
Tel: 1-831-205-9740

Re: [Bloat] buffer sizing meeting notes

[Simon Barber]

The Netflix presentation requires a password - can you share it?

Simon


> On Aug 23, 2019, at 9:07 AM, Dave Taht <dave.taht@gmail.com> wrote:
> 
> There's some good preso from various representatives in the dc market here.
> 
> The p4 stuff, in particular (from barefoot) is looking impressive.
> 
> http://yuba.stanford.edu/~bspang/buffer-sizing-meeting/
> 
> 
> -- 
> 
> Dave Täht
> CTO, TekLibre, LLC
> http://www.teklibre.com
> Tel: 1-831-205-9740
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat

Re: [Bloat] buffer sizing meeting notes

[Dave Taht]

Simon Barber <simon@superduper.net> writes:

> The Netflix presentation requires a password - can you share it?

I was not at this meeting, I just stumbled across the url. I don't even
know how, I think I was looking for the old (and good!) stanford
buffersizing paper, and there was this set of updates from various
parties on the current state of some hardware.


>
> Simon
>
>
>> On Aug 23, 2019, at 9:07 AM, Dave Taht <dave.taht@gmail.com> wrote:
>> 
>> There's some good preso from various representatives in the dc
>> market here.
>> 
>> The p4 stuff, in particular (from barefoot) is looking impressive.
>> 
>> http://yuba.stanford.edu/~bspang/buffer-sizing-meeting/
>> 
>> 
>> -- 
>> 
>> Dave Täht
>> CTO, TekLibre, LLC
>> http://www.teklibre.com
>> Tel: 1-831-205-9740
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat

Re: [Bloat] buffer sizing meeting notes

[Jesper Dangaard Brouer]

On Fri, 23 Aug 2019 10:13:39 -0700
Simon Barber <simon@superduper.net> wrote:

> > On Aug 23, 2019, at 9:07 AM, Dave Taht <dave.taht@gmail.com> wrote:
> > 
> > There's some good preso from various representatives in the dc market here.
> > 
> > The p4 stuff, in particular (from barefoot) is looking impressive.
> > 
> > http://yuba.stanford.edu/~bspang/buffer-sizing-meeting/

I followed the link to the animation of AIMD (Additive
Increase/Multiplicative Decrease), that was really cool!

 https://dabh.github.io/network_animations/

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  LinkedIn: http://www.linkedin.com/in/brouer

Re: [Bloat] buffer sizing meeting notes

[Sebastian Moeller]

Another nugget from the notes (http://yuba.stanford.edu/~bspang/buffer-sizing-meeting/notes/):

Chuanxiong Guo, Bytedance

Talked about buffering in Bytedance’s datacenters. The top of rack switches have buffers of 12-32MB, and the aggregation switches have larger buffers of several GB. Were hoping to use ECN to reduce buffers for aggregation switches, but argued that precise ECN marking on egress queue length in VoQs is hard. However, Arista and Barefoot said they’ve fixed this using the packet latency instead of queue length for ECN.

This looks like an argument for fq_codel/cake's use of time instead of queue length, OR an argument for fq, because in a non-overwhelmed fq_system the local bucket's queue length should be somewhat stronger correlated with the sojurn times, than the sojurn time of a packet though a shared queue, no?

Best Regards
	Sebastian 

> On Aug 24, 2019, at 09:54, Jesper Dangaard Brouer <brouer@redhat.com> wrote:
> 
> On Fri, 23 Aug 2019 10:13:39 -0700
> Simon Barber <simon@superduper.net> wrote:
> 
>>> On Aug 23, 2019, at 9:07 AM, Dave Taht <dave.taht@gmail.com> wrote:
>>> 
>>> There's some good preso from various representatives in the dc market here.
>>> 
>>> The p4 stuff, in particular (from barefoot) is looking impressive.
>>> 
>>> http://yuba.stanford.edu/~bspang/buffer-sizing-meeting/
> 
> I followed the link to the animation of AIMD (Additive
> Increase/Multiplicative Decrease), that was really cool!
> 
> https://dabh.github.io/network_animations/
> 
> -- 
> Best regards,
>  Jesper Dangaard Brouer
>  MSc.CS, Principal Kernel Engineer at Red Hat
>  LinkedIn: http://www.linkedin.com/in/brouer
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat

Re: [Bloat] buffer sizing meeting notes

[Luca Muscariello]

[-- Attachment #1: Type: text/plain, Size: 1474 bytes --]

On Sat, Aug 24, 2019 at 9:59 AM Sebastian Moeller <moeller0@gmx.de> wrote:

> Another nugget from the notes (
> http://yuba.stanford.edu/~bspang/buffer-sizing-meeting/notes/):
>
> This looks like an argument for fq_codel/cake's use of time instead of
> queue length, OR an argument for fq, because in a non-overwhelmed fq_system
> the local bucket's queue length should be somewhat stronger correlated with
> the sojurn times, than the sojurn time of a packet though a shared queue,
> no?
>

I believe that using fq there are good implications in terms of buffer
sizing rules.
The focus is only on backlogged flows. The sparse flow queue can be sized
purely based on queue load.
Which is trivial queuing theory formulas with non-TCP traffic.

For backlogged flows, sizing is just like sizing a buffer for a single TCP
flow. So the computation
can be done based on estimations of the number of backlogged flows and by
setting the cut-off BDP based on the
largest (maximum) min RTT one want to serve optimally.  So AQM is really
needed to get optimality.
The number of backlogged flows could be estimated as a max value or an avg
value, and this of course
changes depending on the network segment (DC, residential, campus, access,
backhaul etc.).

Nick McKeown has done quite a lot of research on the topic (reported in the
slide deck), so I find hilarious the following in his slide deck
"Personal confession: I have no idea what the general answer is" about
sizing buffers!

[-- Attachment #2: Type: text/html, Size: 2064 bytes --]