From mboxrd@z Thu Jan 1 00:00:00 1970
Return-Path: =
I don't doubt that they test. My point was different - there are too =
many knobs and too big a parameter space to test efectively. And that=
's the point. I realize that it's extremely fun to invent parame=
ters in "standards organizations" like 3GPP. Everybody has their own =
favorite knob, and a great rationale for some unusual, but critically "impo=
rtant" customer requirement that might come up some day. Hell, Linux =
has a gazillion (yes, that's a technical term in mathematics!) parameters, =
almost none of which are touched. This reflects the fact that nothing=
ever gets removed once added. LTE is now going into release 12, and =
it's completely ramified into "solutions" to problems that will never be fi=
xed in the field with those solutions. It's great for European Public=
ally Funded Academic-Industry research - lots for those "Professors" to cla=
im they invented. I've worked with telco contractors in the field=
. They don't read manuals, and they don't read specs. The=
y have a job to do, and so much money to spend, and time's a wasting. =
They don't even work for Verizon or ATT. They follow "specs" handed =
down, and charge more if you tell them that the specs have changed.
This is not how brand-new systems get tuned.
=0A=0A
It's a Clown Circ= us out there, and more parameters don't help.
=0A=0A
This is why "more b= uffering is better" continues to be the law of the land - the spec is defin= ed to be "no lost packets under load". I'm sure that the primar= y measure under load for RRUL will be "no lost packets" by the time it gets= to field engineers in the form of "specs" - because that's what they've *a= lways* been told, and they will disregard any changes as "typos".
=0A=0A
A system with more than two control parameters that interact in complex wa= ys is ungovernable - and no control parameters in LTE are "orthogonal", muc= h less "linear" in their interaction.
=0A=0A
=0A
=0A
-----Origi=
nal Message-----
From: "Jim Gettys" <jg@freedesktop.org>
Se=
nt: Friday, March 1, 2013 11:09am
To: "David P Reed" <dpreed@reed.c=
om>
Cc: "Ketan Kulkarni" <ketkulka@gmail.com>, "cerowrt-devel=
@lists.bufferbloat.net" <cerowrt-devel@lists.bufferbloat.net>
Su=
bject: Re: [Cerowrt-devel] Google working on experimental 3.8 Linux kernel =
for Android
=0AOne wonders why all this complexity is necessary, and how likely it= is to be "well tuned" by operators and their contract installers.
=0A=0A
I'm willing to bet $1000 tha= t all the testing that is done is "Can you hear me now" and a "speed test".= Not even something as simple and effective as RRUL.
=0A=0AActually, at least some the the carriers do much more ext= ensive testing; but not with the test tools we would like to see used (yet)= .=0AAn example is AT&T, where in research, KK Ramakrishnan h= as a van with 20 or so laptops so he can go driving around and load up a ce= ll in the middle of the night and get data. And he's research; the o= perations guys do lots of testing I gather, but more at the radio level.=0A=0ANext up, is to educate KK to run RRUL.=0AAnd in my ow= n company, I've seen data, but it is too high level: e.g. performance of "w= eb" video: e.g. siverlight, flash, youtube, etc.=0AA common dise= ase that has complicated all this is the propensity for companies to use Wi= ndows XP internally for everything: since window scaling is turned off, you= can't saturate a LTE link the way you might like to do with a single TCP c= onnection.=0A- Jim=0A=0A=0A=0A=0A=0A=0A&nb= sp;
=0A-----Origi= nal Message-----
=0A
From: "Ketan Kulkarni" <ketkulka@gmail.com>
Sent: Friday,= March 1, 2013 3:00am
To: "Jim Gettys" <jg@freedesktop.org>
Cc: "cerowrt-d= evel@lists.bufferbloat.net" <cerowrt-devel@lists.bufferbloat.net&g= t;
Subject: Re: [Cerowrt-devel] Google working on experimental 3.8 Li= nux kernel for Android=0A=0A=0A
=0AOn Fri, Mar 1, 2013 at 1:33 AM, Jim Gettys &= lt;jg@freedesktop.o= rg> wrote:=0A=0A
=0A=0AI've got a bit more insight into LTE than I did in the past,= courtesy of the last couple days.=0A To begin with, LTE runs with seve= ral classes of service (the call them bearers). Your VOIP traffic goe= s into one of them.=0AAnd I think there is another as well that = is for guaranteed bit rate traffic. One transmit opportunity may have= a bunch of chunks of data, and that data may be destined for more than one= device (IIRC). It's substantially different than WiFi.=0AJust thought to put more light on bearer stuff:=0A
There are two ways bearers are setup:
1. UE initiated - wher= e User Equipment sets-up the "parameters" for bearer
2. Network init= iated - where node like PCRF and PGW sets-up the "parameters".
Param= eters include the Guaranteed bit-rates, maximum bit-rates. Something called= QCI is associated with bearers. The QCI parameters are authorized at PCRF = (policy control rule function) and there is certain mapping maintained at e= ither PCRF or PGW between QCI values and DSCP and MBRs.
These paramet= ers enforcing is done at PGW (in such case it is termed as PCEF - policy an= d rule enforcement function). So PGWs depending on bearers can certainly mo= dify dscp bits. Though these can be modified by other nodes in the network.=
There are two types of bearers: 1. Dedicated bearers - to carr= y traffic which need "special" treatment 2. Default or general pupose beare= rs - to carry all general purpose data.
So generally the voip, streami= ng videos are passed over dedicated bearers and apply (generally) higher GB= Rs, MBRs and correct dscp markings.
And other non-latency sensitive t= raffic generally follows the default bearer.
Theoretical limit o= n maximum bearers is 11 though practically most of the deployments use upto= 3 bearers max.
Note that these parameters may very well very ba= sed on the subscriber profiles. Premium/Corporate subscribers can well have= more GBRs and MBRs.
ISPs are generally very much sensitive to the co= rrect markings at gateways for obvious reasons.=0A =0A=0ABut most of what = we think of as Internet stuff (web surfing, dns, etc) all gets dumped into = a single best effort ("BE"), class.=0AThe BE class is defin= itely badly bloated; I can't say how much because I don't really know = yet; the test my colleague ran wasn't run long enough to be confident it fi= lled the buffers). But I will say worse than most cable modems I've s= een. I expect this will be true to different degrees on different har= dware. The other traffic classes haven't been tested yet for bufferbl= oat, though I suspect they will have it too. I was told that those cl= asses have much shorter queues, and when the grow, they dump the whole queu= es (because delivering late real time traffic is useless). But trust = *and* verify.... Verification hasn't been done for anything but BE tr= affic, and that hasn't been quantified.=0ABut each device gets a= "fair" shot at bandwidth in the cell (or sector of a cell; they run 3 radi= os in each cell), where fair is basically time based; if you are at the edg= e of a cell, you'll get a lot less bandwidth than someone near a tower; and= this fairness is guaranteed by a scheduler than runs in the base station (= called a b-nodeb, IIIRC). So the base station guarantees some sort of= "fairness" between devices (a place where Linux's wifi stack today fails u= tterly, since there is a single queue per device, rather than one per stati= on).=0AWhether there are bloat problems at the link level in LTE= due to error correction I don't know yet; but it wouldn't surprise me; I k= now there was in 3g. The people I talked to this morning aren't famil= iar with the HARQ layer in the system.=0AThe base stations are c= omplicated beasts; they have both a linux system in them as well as a real = time operating system based device inside We don't know where the bot= tle neck(s) are yet. I spent lunch upping their paranoia and getting = them through some conceptual hurdles (e.g. multiple bottlenecks that may mo= ve, and the like). They will try to get me some of the data so I can = help them figure it out. I don't know if the data flow goes through t= he linux system in the bnodeb or not, for example.=0AMost carrie= rs are now trying to ensure that their backhauls from the base station are = never congested, though that is another known source of problems. And= then there is the lack of AQM at peering point routers.... You'd thi= nk they might run WRED there, but many/most do not.=0A=0A- Jim=0A= =0A=0A=0A=0A=0A
=0AOn Thu, Feb 28, 2013 at 2:08 PM, Dave Taht <dave.taht@gmail.com= > wrote:=0A
=0A
=0A=0A=0AOn Thu, Feb 28, 2013 at 1:57 PM,=0A<dp= reed@reed.com> wrote:
=0A=0A=0ADoesn't fq_codel need an estimate of lin= k capacity?
=0ANo, it just measures= delay. Since so far as I know the outgoing portion of LTE is not soft-rate= limited, but sensitive to the actual available link bandwidth, fq_codel sh= ould work pretty good (if the underlying interfaces weren't horribly overbu= ffired) in that direction.=0AI'm looking forward to some measure= ments of actual buffering at the device driver/device levels.=0A= I don't know how inbound to the handset is managed via LTE.=0ASt= ill quite a few assumptions left to smash in the above.=0A...=0A=0Ain the home router case....=0A...=0AWhen t= here are artificial rate limits in play (in, for example, a cable modem/CMT= S, hooked up via gigE yet rate limiting to 24up/4mbit down), then a rate li= miter (tbf,htb,hfsc) needs to be applied locally to move that rate limiter/= queue management into the local device, se we can manage it better.= =0AI'd like to be rid of the need to use htb and come up with a rate l= imiter that could be adjusted dynamically from a daemon in userspace, probi= ng for short all bandwidth fluctuations while monitoring the load. It neede= nt send that much data very often, to come up with a stable result....=0AYou've described one soft-rate sensing scheme (piggybacking on TCP= ), and I've thought up a few others, that could feed back from a daemon som= e samples into a a soft(er) rate limiter that would keep control of the que= ues in the home router. I am thinking it's going to take way too long to fi= x the CPE and far easier to fix the home router via this method, and certai= nly it's too painful and inaccurate to merely measure the bandwidth once, t= hen set a hard rate, when=0ASo far as I know the gargoyle projec= t was experimenting with this approach.=0AA problem is in places= that connect more than one device to the cable modem... then you end up wi= th those needing to communicate their perception of the actual bandwidth be= yond the link.=0A=0A=0A=0A=0A=0A____________________________________= ___________Where will it get that from the 4G or 3G upli= nk?
=0A=0A=0A=0A=0A=0A
-= ----Original Message-----
=0A
From: "Maciej Soltysiak" <maciej@soltysiak.com>
Sent: Thursday, February 28, 2013 1:03pm
To: cerowrt-devel@lists.buff= erbloat.net
Subject: [Cerowrt-devel] Google working on experimenta= l 3.8 Linux kernel for Android=0A=0A=0AHiya,= =0ALooks like Google's experimenting with 3.8 for Android: https://android.googlesource.com/kernel/common/+/experime= ntal/android-3.8=0ASounds great if this means they will util= ize fq_codel, TFO, BQL, etc.=0AAnyway my nexus 7 says it has 3.1= .10 and this 3.8 will probably go to Android 5.0 so I hope Nexus 7 will get= it too some day or at least 3.3+=0APhoronix coverage: http://www.phoronix.com/scan.php?page=3Dnews_item&px= =3DMTMxMzc=0ATheir 3.8 changelog: https://android.googlesource.com/kernel/common/+log/experimental/andr= oid-3.8=0ARegards,=0AMaciej=0A
Cerowrt-devel mailing list
Cerowrt-devel@lists.bufferb= loat.net
https://lists.bufferbloat.net/listinfo/cerowrt-de= vel
=0A
--
Dave T=C3=A4ht
Fixing bufferbloat wi= th cerowrt: http://www.teklibre.com/cerowrt/subscribe.html
_______________________________________________
Cerowrt-de= vel mailing list
Cerowrt-devel@lists.bufferbloat.net
ht= tps://lists.bufferbloat.net/listinfo/cerowrt-devel
=0A
__________________________= _____________________
Cerowrt-devel mailing list
Cerowrt-devel@lis= ts.bufferbloat.net
https://lists.bufferbloat.net/listinfo/= cerowrt-devel
=0A