Development issues regarding the cerowrt test router project
 help / color / mirror / Atom feed
* [Cerowrt-devel] Routing limit question
@ 2014-10-19 17:51 Ernesto Elias
  2014-10-19 18:24 ` Dave Taht
  0 siblings, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-19 17:51 UTC (permalink / raw)
  To: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 445 bytes --]

Hello everyone!
I have a question about the wndr3800 routing limit. I went back to the
older submissions to see if I can find what would be the answer for it. But
in my search I haven't managed to find a definite answer. From what I seen
about setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
wondering if anyone can shed some light for me here as I have verizon fios
and my speeds are 50 dl/50 ul. Thank you guys very much!

[-- Attachment #2: Type: text/html, Size: 485 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 17:51 [Cerowrt-devel] Routing limit question Ernesto Elias
@ 2014-10-19 18:24 ` Dave Taht
  2014-10-19 18:48   ` Ernesto Elias
  2014-10-19 18:56   ` Sebastian Moeller
  0 siblings, 2 replies; 24+ messages in thread
From: Dave Taht @ 2014-10-19 18:24 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

On at least one verizon device I've tried it appeared that they had
SFQ or something similar on egress from the modem.

http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down

So you only needed to shape the download. which is good as we start
peaking out at 50Mbit download total. But only measurements can tell.


On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> Hello everyone!
> I have a question about the wndr3800 routing limit. I went back to the older
> submissions to see if I can find what would be the answer for it. But in my
> search I haven't managed to find a definite answer. From what I seen about
> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> wondering if anyone can shed some light for me here as I have verizon fios
> and my speeds are 50 dl/50 ul. Thank you guys very much!
>
> _______________________________________________
> Cerowrt-devel mailing list
> Cerowrt-devel@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>



-- 
Dave Täht

thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 18:24 ` Dave Taht
@ 2014-10-19 18:48   ` Ernesto Elias
  2014-10-19 18:51     ` Dave Taht
  2014-10-19 18:56   ` Sebastian Moeller
  1 sibling, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-19 18:48 UTC (permalink / raw)
  To: Dave Taht; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 1640 bytes --]

Oh OK I mean I was just wondering because I was trying to figure out what
should I have put in the sqm tab because I saw that the wndr3800 tops out
at 50 mbit combined of download and upload right? As of now I have it at
25/25 on the page. Is that right or should I just shape the upload?
On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:

> On at least one verizon device I've tried it appeared that they had
> SFQ or something similar on egress from the modem.
>
>
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
>
> So you only needed to shape the download. which is good as we start
> peaking out at 50Mbit download total. But only measurements can tell.
>
>
> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> > Hello everyone!
> > I have a question about the wndr3800 routing limit. I went back to the
> older
> > submissions to see if I can find what would be the answer for it. But in
> my
> > search I haven't managed to find a definite answer. From what I seen
> about
> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> > wondering if anyone can shed some light for me here as I have verizon
> fios
> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> >
> > _______________________________________________
> > Cerowrt-devel mailing list
> > Cerowrt-devel@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> >
>
>
>
> --
> Dave Täht
>
> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>

[-- Attachment #2: Type: text/html, Size: 2407 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 18:48   ` Ernesto Elias
@ 2014-10-19 18:51     ` Dave Taht
  2014-10-19 19:16       ` Ernesto Elias
  0 siblings, 1 reply; 24+ messages in thread
From: Dave Taht @ 2014-10-19 18:51 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

shape the download only, then measure. shape the upload only, then measure.

On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> Oh OK I mean I was just wondering because I was trying to figure out what
> should I have put in the sqm tab because I saw that the wndr3800 tops out at
> 50 mbit combined of download and upload right? As of now I have it at 25/25
> on the page. Is that right or should I just shape the upload?
>
> On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
>>
>> On at least one verizon device I've tried it appeared that they had
>> SFQ or something similar on egress from the modem.
>>
>>
>> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
>>
>> So you only needed to shape the download. which is good as we start
>> peaking out at 50Mbit download total. But only measurements can tell.
>>
>>
>> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
>> wrote:
>> > Hello everyone!
>> > I have a question about the wndr3800 routing limit. I went back to the
>> > older
>> > submissions to see if I can find what would be the answer for it. But in
>> > my
>> > search I haven't managed to find a definite answer. From what I seen
>> > about
>> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
>> > wondering if anyone can shed some light for me here as I have verizon
>> > fios
>> > and my speeds are 50 dl/50 ul. Thank you guys very much!
>> >
>> > _______________________________________________
>> > Cerowrt-devel mailing list
>> > Cerowrt-devel@lists.bufferbloat.net
>> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
>> >
>>
>>
>>
>> --
>> Dave Täht
>>
>> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks



-- 
Dave Täht

thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 18:24 ` Dave Taht
  2014-10-19 18:48   ` Ernesto Elias
@ 2014-10-19 18:56   ` Sebastian Moeller
  2014-10-19 19:22     ` Dave Taht
       [not found]     ` <CAA93jw6DFxoDaRVCoQn=zxbOjgQ-dA1bCZ1W=p9br25jWrRgkg@mail.gmail.com>
  1 sibling, 2 replies; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-19 18:56 UTC (permalink / raw)
  To: Dave Täht; +Cc: Ernesto Elias, cerowrt-devel

HI Dave,


On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:

> On at least one verizon device I've tried it appeared that they had
> SFQ or something similar on egress from the modem.
> 
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> 
> So you only needed to shape the download. which is good as we start
> peaking out at 50Mbit download total. But only measurements can tell.

	So on Hnymans community openwrt build a few fortunate ones on excellent lines seem to get decent results even at 110-120 Mbps combined:
https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
and:
https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
I have no idea why and both lines were reasonably well-behaved even without any AQM/QOS...

Also I wonder whether when we increase the quantum for higher rates to give HTB some breathing room, whether we also should increase burst and cburst? My hunch is that quantum affects the switching between the leaves, while busts and cburst should allow to dump more data to lower layers inside each leaf qdisc. And since we are running behind, maybe taking a bigger shovel can help some. (I assume this needs to be titrated not to kill latency under load, but if we can only effective have HTB execute x times per second we can easily afford to dump line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet link is way to slow to test this...

Best Regards
	Sebastian

> 
> 
> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
>> Hello everyone!
>> I have a question about the wndr3800 routing limit. I went back to the older
>> submissions to see if I can find what would be the answer for it. But in my
>> search I haven't managed to find a definite answer. From what I seen about
>> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
>> wondering if anyone can shed some light for me here as I have verizon fios
>> and my speeds are 50 dl/50 ul. Thank you guys very much!
>> 
>> _______________________________________________
>> Cerowrt-devel mailing list
>> Cerowrt-devel@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>> 
> 
> 
> 
> -- 
> Dave Täht
> 
> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> _______________________________________________
> Cerowrt-devel mailing list
> Cerowrt-devel@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cerowrt-devel


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 18:51     ` Dave Taht
@ 2014-10-19 19:16       ` Ernesto Elias
  2014-10-19 19:55         ` Sebastian Moeller
  0 siblings, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-19 19:16 UTC (permalink / raw)
  To: Dave Taht; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 3305 bytes --]

root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4 streams
down a

            nd up while pinging gstatic.com. Takes about 60 seconds.
 Download:  30.49 Mbps
   Upload:  46.52 Mbps
  Latency: (in msec, 61 pings, 0.00% packet loss)
      Min: 19.848
    10pct: 19.958
   Median: 20.743
      Avg: 21.229
    90pct: 22.739
      Max: 30.491
root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4 streams
down a

            nd up while pinging gstatic.com. Takes about 60 seconds.
 Download:  7.02 Mbps
   Upload:  63.68 Mbps
  Latency: (in msec, 61 pings, 0.00% packet loss)
      Min: 21.690
    10pct: 117.997
   Median: 153.983
      Avg: 150.851
    90pct: 180.777
      Max: 190.265


The first result i gotten by only shaping the uplink and the second results
are for only shaping the downlink...

and for the testing i put 0 on the for the downlink and uplink when I was
testing for the opposite. and I put 50000 in it when I was shaping it for
the test. Am I doing it right ?

On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com> wrote:

> shape the download only, then measure. shape the upload only, then measure.
>
> On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> > Oh OK I mean I was just wondering because I was trying to figure out what
> > should I have put in the sqm tab because I saw that the wndr3800 tops
> out at
> > 50 mbit combined of download and upload right? As of now I have it at
> 25/25
> > on the page. Is that right or should I just shape the upload?
> >
> > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> >>
> >> On at least one verizon device I've tried it appeared that they had
> >> SFQ or something similar on egress from the modem.
> >>
> >>
> >>
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> >>
> >> So you only needed to shape the download. which is good as we start
> >> peaking out at 50Mbit download total. But only measurements can tell.
> >>
> >>
> >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <
> ernestogelias@gmail.com>
> >> wrote:
> >> > Hello everyone!
> >> > I have a question about the wndr3800 routing limit. I went back to the
> >> > older
> >> > submissions to see if I can find what would be the answer for it. But
> in
> >> > my
> >> > search I haven't managed to find a definite answer. From what I seen
> >> > about
> >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> >> > wondering if anyone can shed some light for me here as I have verizon
> >> > fios
> >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> >> >
> >> > _______________________________________________
> >> > Cerowrt-devel mailing list
> >> > Cerowrt-devel@lists.bufferbloat.net
> >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> >> >
> >>
> >>
> >>
> >> --
> >> Dave Täht
> >>
> >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>
>
>
> --
> Dave Täht
>
> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>

[-- Attachment #2: Type: text/html, Size: 5553 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 18:56   ` Sebastian Moeller
@ 2014-10-19 19:22     ` Dave Taht
  2014-10-19 19:33       ` Ernesto Elias
       [not found]     ` <CAA93jw6DFxoDaRVCoQn=zxbOjgQ-dA1bCZ1W=p9br25jWrRgkg@mail.gmail.com>
  1 sibling, 1 reply; 24+ messages in thread
From: Dave Taht @ 2014-10-19 19:22 UTC (permalink / raw)
  To: Sebastian Moeller; +Cc: Ernesto Elias, cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 2868 bytes --]

It looks to me as if the actiontec has pretty good qos all by itself.
On Oct 19, 2014 11:56 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:

> HI Dave,
>
>
> On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:
>
> > On at least one verizon device I've tried it appeared that they had
> > SFQ or something similar on egress from the modem.
> >
> >
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> >
> > So you only needed to shape the download. which is good as we start
> > peaking out at 50Mbit download total. But only measurements can tell.
>
>         So on Hnymans community openwrt build a few fortunate ones on
> excellent lines seem to get decent results even at 110-120 Mbps combined:
> https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
> and:
> https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
> I have no idea why and both lines were reasonably well-behaved even
> without any AQM/QOS...
>
> Also I wonder whether when we increase the quantum for higher rates to
> give HTB some breathing room, whether we also should increase burst and
> cburst? My hunch is that quantum affects the switching between the leaves,
> while busts and cburst should allow to dump more data to lower layers
> inside each leaf qdisc. And since we are running behind, maybe taking a
> bigger shovel can help some. (I assume this needs to be titrated not to
> kill latency under load, but if we can only effective have HTB execute x
> times per second we can easily afford to dump
> line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet
> link is way to slow to test this...
>
> Best Regards
>         Sebastian
>
> >
> >
> > On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> >> Hello everyone!
> >> I have a question about the wndr3800 routing limit. I went back to the
> older
> >> submissions to see if I can find what would be the answer for it. But
> in my
> >> search I haven't managed to find a definite answer. From what I seen
> about
> >> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> >> wondering if anyone can shed some light for me here as I have verizon
> fios
> >> and my speeds are 50 dl/50 ul. Thank you guys very much!
> >>
> >> _______________________________________________
> >> Cerowrt-devel mailing list
> >> Cerowrt-devel@lists.bufferbloat.net
> >> https://lists.bufferbloat.net/listinfo/cerowrt-devel
> >>
> >
> >
> >
> > --
> > Dave Täht
> >
> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > _______________________________________________
> > Cerowrt-devel mailing list
> > Cerowrt-devel@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
>
>

[-- Attachment #2: Type: text/html, Size: 4119 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 19:22     ` Dave Taht
@ 2014-10-19 19:33       ` Ernesto Elias
  2014-10-19 19:41         ` Dave Taht
  0 siblings, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-19 19:33 UTC (permalink / raw)
  To: Dave Taht; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 3073 bytes --]

I'm not using the actiontec at all. I'm using the wndr3800
On Oct 19, 2014 3:22 PM, "Dave Taht" <dave.taht@gmail.com> wrote:

> It looks to me as if the actiontec has pretty good qos all by itself.
> On Oct 19, 2014 11:56 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
>
>> HI Dave,
>>
>>
>> On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:
>>
>> > On at least one verizon device I've tried it appeared that they had
>> > SFQ or something similar on egress from the modem.
>> >
>> >
>> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
>> >
>> > So you only needed to shape the download. which is good as we start
>> > peaking out at 50Mbit download total. But only measurements can tell.
>>
>>         So on Hnymans community openwrt build a few fortunate ones on
>> excellent lines seem to get decent results even at 110-120 Mbps combined:
>> https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
>> and:
>> https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
>> I have no idea why and both lines were reasonably well-behaved even
>> without any AQM/QOS...
>>
>> Also I wonder whether when we increase the quantum for higher rates to
>> give HTB some breathing room, whether we also should increase burst and
>> cburst? My hunch is that quantum affects the switching between the leaves,
>> while busts and cburst should allow to dump more data to lower layers
>> inside each leaf qdisc. And since we are running behind, maybe taking a
>> bigger shovel can help some. (I assume this needs to be titrated not to
>> kill latency under load, but if we can only effective have HTB execute x
>> times per second we can easily afford to dump
>> line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet
>> link is way to slow to test this...
>>
>> Best Regards
>>         Sebastian
>>
>> >
>> >
>> > On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <
>> ernestogelias@gmail.com> wrote:
>> >> Hello everyone!
>> >> I have a question about the wndr3800 routing limit. I went back to the
>> older
>> >> submissions to see if I can find what would be the answer for it. But
>> in my
>> >> search I haven't managed to find a definite answer. From what I seen
>> about
>> >> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
>> >> wondering if anyone can shed some light for me here as I have verizon
>> fios
>> >> and my speeds are 50 dl/50 ul. Thank you guys very much!
>> >>
>> >> _______________________________________________
>> >> Cerowrt-devel mailing list
>> >> Cerowrt-devel@lists.bufferbloat.net
>> >> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>> >>
>> >
>> >
>> >
>> > --
>> > Dave Täht
>> >
>> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>> > _______________________________________________
>> > Cerowrt-devel mailing list
>> > Cerowrt-devel@lists.bufferbloat.net
>> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>
>>

[-- Attachment #2: Type: text/html, Size: 4577 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 19:33       ` Ernesto Elias
@ 2014-10-19 19:41         ` Dave Taht
  0 siblings, 0 replies; 24+ messages in thread
From: Dave Taht @ 2014-10-19 19:41 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 3424 bytes --]

Sorry there are two conversational threads here on hynyman s thread they
are benchmarking vs an anctiontec.

Also it does seem like fiddling with the burst parameter to htb seems
warrented.
On Oct 19, 2014 12:33 PM, "Ernesto Elias" <ernestogelias@gmail.com> wrote:

> I'm not using the actiontec at all. I'm using the wndr3800
> On Oct 19, 2014 3:22 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
>
>> It looks to me as if the actiontec has pretty good qos all by itself.
>> On Oct 19, 2014 11:56 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
>>
>>> HI Dave,
>>>
>>>
>>> On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:
>>>
>>> > On at least one verizon device I've tried it appeared that they had
>>> > SFQ or something similar on egress from the modem.
>>> >
>>> >
>>> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
>>> >
>>> > So you only needed to shape the download. which is good as we start
>>> > peaking out at 50Mbit download total. But only measurements can tell.
>>>
>>>         So on Hnymans community openwrt build a few fortunate ones on
>>> excellent lines seem to get decent results even at 110-120 Mbps combined:
>>> https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
>>> and:
>>> https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
>>> I have no idea why and both lines were reasonably well-behaved even
>>> without any AQM/QOS...
>>>
>>> Also I wonder whether when we increase the quantum for higher rates to
>>> give HTB some breathing room, whether we also should increase burst and
>>> cburst? My hunch is that quantum affects the switching between the leaves,
>>> while busts and cburst should allow to dump more data to lower layers
>>> inside each leaf qdisc. And since we are running behind, maybe taking a
>>> bigger shovel can help some. (I assume this needs to be titrated not to
>>> kill latency under load, but if we can only effective have HTB execute x
>>> times per second we can easily afford to dump
>>> line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet
>>> link is way to slow to test this...
>>>
>>> Best Regards
>>>         Sebastian
>>>
>>> >
>>> >
>>> > On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <
>>> ernestogelias@gmail.com> wrote:
>>> >> Hello everyone!
>>> >> I have a question about the wndr3800 routing limit. I went back to
>>> the older
>>> >> submissions to see if I can find what would be the answer for it. But
>>> in my
>>> >> search I haven't managed to find a definite answer. From what I seen
>>> about
>>> >> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
>>> >> wondering if anyone can shed some light for me here as I have verizon
>>> fios
>>> >> and my speeds are 50 dl/50 ul. Thank you guys very much!
>>> >>
>>> >> _______________________________________________
>>> >> Cerowrt-devel mailing list
>>> >> Cerowrt-devel@lists.bufferbloat.net
>>> >> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>> >>
>>> >
>>> >
>>> >
>>> > --
>>> > Dave Täht
>>> >
>>> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>>> > _______________________________________________
>>> > Cerowrt-devel mailing list
>>> > Cerowrt-devel@lists.bufferbloat.net
>>> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>>
>>>

[-- Attachment #2: Type: text/html, Size: 5124 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 19:16       ` Ernesto Elias
@ 2014-10-19 19:55         ` Sebastian Moeller
  2014-10-20  1:08           ` Ernesto Elias
  0 siblings, 1 reply; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-19 19:55 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

Hi Ernesto,

I would recommend to not run the netperf processes on the router you want to assess, if possible. Shaping at the rates you are interested in probably will take 100% CPU of the router, so there is no slack for running net server and netperf instances at the same time. Also I would recommend to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper) instead of netperfrunner.sh as it allows so much nicer visualization of the queueing over time. netperfrunner.sh is great tool for quick testing and figuring out link capacities and worst case latency under load increase (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper will run under linux and macosx natively, but will also run (well enough) in a linux virtual machine on a windows box.


Best Regards
	Sebastian


On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com> wrote:

> root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
>  Download:  30.49 Mbps
>    Upload:  46.52 Mbps
>   Latency: (in msec, 61 pings, 0.00% packet loss)
>       Min: 19.848
>     10pct: 19.958
>    Median: 20.743
>       Avg: 21.229
>     90pct: 22.739
>       Max: 30.491
> root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
>  Download:  7.02 Mbps
>    Upload:  63.68 Mbps
>   Latency: (in msec, 61 pings, 0.00% packet loss)
>       Min: 21.690
>     10pct: 117.997
>    Median: 153.983
>       Avg: 150.851
>     90pct: 180.777
>       Max: 190.265
> 
> 
> The first result i gotten by only shaping the uplink and the second results are for only shaping the downlink...
> 
> and for the testing i put 0 on the for the downlink and uplink when I was testing for the opposite. and I put 50000 in it when I was shaping it for the test. Am I doing it right ?
> 
> On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com> wrote:
> shape the download only, then measure. shape the upload only, then measure.
> 
> On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> > Oh OK I mean I was just wondering because I was trying to figure out what
> > should I have put in the sqm tab because I saw that the wndr3800 tops out at
> > 50 mbit combined of download and upload right? As of now I have it at 25/25
> > on the page. Is that right or should I just shape the upload?
> >
> > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> >>
> >> On at least one verizon device I've tried it appeared that they had
> >> SFQ or something similar on egress from the modem.
> >>
> >>
> >> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> >>
> >> So you only needed to shape the download. which is good as we start
> >> peaking out at 50Mbit download total. But only measurements can tell.
> >>
> >>
> >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
> >> wrote:
> >> > Hello everyone!
> >> > I have a question about the wndr3800 routing limit. I went back to the
> >> > older
> >> > submissions to see if I can find what would be the answer for it. But in
> >> > my
> >> > search I haven't managed to find a definite answer. From what I seen
> >> > about
> >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> >> > wondering if anyone can shed some light for me here as I have verizon
> >> > fios
> >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> >> >
> >> > _______________________________________________
> >> > Cerowrt-devel mailing list
> >> > Cerowrt-devel@lists.bufferbloat.net
> >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> >> >
> >>
> >>
> >>
> >> --
> >> Dave Täht
> >>
> >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> 
> 
> 
> --
> Dave Täht
> 
> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> 
> _______________________________________________
> Cerowrt-devel mailing list
> Cerowrt-devel@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cerowrt-devel


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
       [not found]     ` <CAA93jw6DFxoDaRVCoQn=zxbOjgQ-dA1bCZ1W=p9br25jWrRgkg@mail.gmail.com>
@ 2014-10-19 22:44       ` Sebastian Moeller
  2014-10-20 20:04         ` Dave Taht
  0 siblings, 1 reply; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-19 22:44 UTC (permalink / raw)
  To: Dave Täht; +Cc: cerowrt-devel

Hi Dave,

so I just went with what I have available, shaping between linux host on se00 and macbook on sw10, shaper on se00 (25M/25M that is the maximum wireless throughput with the current router position): burst/cburst set at 1600(default) 16000 and 16000. And lo and behold the sirq gets smaller the larger burst/cburst is set. Now tho test is just too confounded by my bad wireless to be proof, but it certainly justifies the time to expose knobs in the GUI to set burst/cburst/quantum values for HTB for each shaper instance independent for ingress and egress… (I do not assume that this will even double the throughput of a wndr as a router, but even just 10-20% will make a difference ;) (I will eat my own dogwood, since I am about to upgrade from 16M/2.5M to 50M/10M right into where our sirq pain starts)).



Best Regards
	Sebastian

On Oct 19, 2014, at 21:27 , Dave Taht <dave.taht@gmail.com> wrote:

> Yes fiddling with burst seems to make sense. Try 16k
> 
> On Oct 19, 2014 11:56 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> HI Dave,
> 
> 
> On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:
> 
> > On at least one verizon device I've tried it appeared that they had
> > SFQ or something similar on egress from the modem.
> >
> > http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> >
> > So you only needed to shape the download. which is good as we start
> > peaking out at 50Mbit download total. But only measurements can tell.
> 
>         So on Hnymans community openwrt build a few fortunate ones on excellent lines seem to get decent results even at 110-120 Mbps combined:
> https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
> and:
> https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
> I have no idea why and both lines were reasonably well-behaved even without any AQM/QOS...
> 
> Also I wonder whether when we increase the quantum for higher rates to give HTB some breathing room, whether we also should increase burst and cburst? My hunch is that quantum affects the switching between the leaves, while busts and cburst should allow to dump more data to lower layers inside each leaf qdisc. And since we are running behind, maybe taking a bigger shovel can help some. (I assume this needs to be titrated not to kill latency under load, but if we can only effective have HTB execute x times per second we can easily afford to dump line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet link is way to slow to test this...
> 
> Best Regards
>         Sebastian
> 
> >
> >
> > On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> >> Hello everyone!
> >> I have a question about the wndr3800 routing limit. I went back to the older
> >> submissions to see if I can find what would be the answer for it. But in my
> >> search I haven't managed to find a definite answer. From what I seen about
> >> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> >> wondering if anyone can shed some light for me here as I have verizon fios
> >> and my speeds are 50 dl/50 ul. Thank you guys very much!
> >>
> >> _______________________________________________
> >> Cerowrt-devel mailing list
> >> Cerowrt-devel@lists.bufferbloat.net
> >> https://lists.bufferbloat.net/listinfo/cerowrt-devel
> >>
> >
> >
> >
> > --
> > Dave Täht
> >
> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > _______________________________________________
> > Cerowrt-devel mailing list
> > Cerowrt-devel@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> 


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 19:55         ` Sebastian Moeller
@ 2014-10-20  1:08           ` Ernesto Elias
  2014-10-20  8:11             ` Sebastian Moeller
  0 siblings, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-20  1:08 UTC (permalink / raw)
  To: Sebastian Moeller; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 5299 bytes --]

Well Sebastian sir, I am taking your advice and I'm on windows 7 so right
now I'm downloading virtualbox and xubuntu. I don't know if there is one
that you guys recommend. Well... I am new to all this stuff so I'm asking
you to please help me on the subject like after everything is up and
running on the computer, what other steps do I do afterwards and forgive me
like I said I'm new ot all this stuff because I want to be able to find my
link capacity when using SQM. and Thanks for the help!

On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de> wrote:

> Hi Ernesto,
>
> I would recommend to not run the netperf processes on the router you want
> to assess, if possible. Shaping at the rates you are interested in probably
> will take 100% CPU of the router, so there is no slack for running net
> server and netperf instances at the same time. Also I would recommend to
> use netperf-wrapper (https://github.com/tohojo/netperf-wrapper) instead
> of netperfrunner.sh as it allows so much nicer visualization of the
> queueing over time. netperfrunner.sh is great tool for quick testing and
> figuring out link capacities and worst case latency under load increase
> (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper
> will run under linux and macosx natively, but will also run (well enough)
> in a linux virtual machine on a windows box.
>
>
> Best Regards
>         Sebastian
>
>
> On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com> wrote:
>
> > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> >  Download:  30.49 Mbps
> >    Upload:  46.52 Mbps
> >   Latency: (in msec, 61 pings, 0.00% packet loss)
> >       Min: 19.848
> >     10pct: 19.958
> >    Median: 20.743
> >       Avg: 21.229
> >     90pct: 22.739
> >       Max: 30.491
> > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> >  Download:  7.02 Mbps
> >    Upload:  63.68 Mbps
> >   Latency: (in msec, 61 pings, 0.00% packet loss)
> >       Min: 21.690
> >     10pct: 117.997
> >    Median: 153.983
> >       Avg: 150.851
> >     90pct: 180.777
> >       Max: 190.265
> >
> >
> > The first result i gotten by only shaping the uplink and the second
> results are for only shaping the downlink...
> >
> > and for the testing i put 0 on the for the downlink and uplink when I
> was testing for the opposite. and I put 50000 in it when I was shaping it
> for the test. Am I doing it right ?
> >
> > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com> wrote:
> > shape the download only, then measure. shape the upload only, then
> measure.
> >
> > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> > > Oh OK I mean I was just wondering because I was trying to figure out
> what
> > > should I have put in the sqm tab because I saw that the wndr3800 tops
> out at
> > > 50 mbit combined of download and upload right? As of now I have it at
> 25/25
> > > on the page. Is that right or should I just shape the upload?
> > >
> > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> > >>
> > >> On at least one verizon device I've tried it appeared that they had
> > >> SFQ or something similar on egress from the modem.
> > >>
> > >>
> > >>
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > >>
> > >> So you only needed to shape the download. which is good as we start
> > >> peaking out at 50Mbit download total. But only measurements can tell.
> > >>
> > >>
> > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <
> ernestogelias@gmail.com>
> > >> wrote:
> > >> > Hello everyone!
> > >> > I have a question about the wndr3800 routing limit. I went back to
> the
> > >> > older
> > >> > submissions to see if I can find what would be the answer for it.
> But in
> > >> > my
> > >> > search I haven't managed to find a definite answer. From what I seen
> > >> > about
> > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> > >> > wondering if anyone can shed some light for me here as I have
> verizon
> > >> > fios
> > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > >> >
> > >> > _______________________________________________
> > >> > Cerowrt-devel mailing list
> > >> > Cerowrt-devel@lists.bufferbloat.net
> > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > >> >
> > >>
> > >>
> > >>
> > >> --
> > >> Dave Täht
> > >>
> > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> >
> >
> >
> > --
> > Dave Täht
> >
> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> >
> > _______________________________________________
> > Cerowrt-devel mailing list
> > Cerowrt-devel@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
>
>

[-- Attachment #2: Type: text/html, Size: 8041 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-20  1:08           ` Ernesto Elias
@ 2014-10-20  8:11             ` Sebastian Moeller
  2014-10-21  2:30               ` Ernesto Elias
  0 siblings, 1 reply; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-20  8:11 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

Hi Ernesto,

On Oct 20, 2014, at 03:08 , Ernesto Elias <ernestogelias@gmail.com> wrote:

> 
> Well Sebastian sir, I am taking your advice and I'm on windows 7 so right now I'm downloading virtualbox and xubuntu. I don't know if there is one that you guys recommend.

	I only tested ubuntu 12.04LTS, but certainly I would try the current 14.07 LTS today.

>  Well... I am new to all this stuff so I'm asking you to please help me on the subject like after everything is up and running on the computer,

	Oh, sure I am happy to help you with getting better data ;)

> what other steps do I do afterwards and forgive me like I said I'm new ot all this stuff because I want to be able to find my link capacity when using SQM. and Thanks for the help!

	First step is getting the vitalized linux up and running, then python 2.7 (I assume that python will be part of the distribution), with matplotlib; and finally netperf-wrapper. I assume it might take a bit trying to figure out all the required packets installed, but it should be reasonably quick to get this up and running… Just give it a try and let me know if/where you get stuck
	It would be most excellent if you could document the whole procedure somehow so we could put it somewhere on the net to make it easier in the future...

Best Regards
	Sebastian

> 
> On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
> Hi Ernesto,
> 
> I would recommend to not run the netperf processes on the router you want to assess, if possible. Shaping at the rates you are interested in probably will take 100% CPU of the router, so there is no slack for running net server and netperf instances at the same time. Also I would recommend to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper) instead of netperfrunner.sh as it allows so much nicer visualization of the queueing over time. netperfrunner.sh is great tool for quick testing and figuring out link capacities and worst case latency under load increase (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper will run under linux and macosx natively, but will also run (well enough) in a linux virtual machine on a windows box.
> 
> 
> Best Regards
>         Sebastian
> 
> 
> On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> 
> > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> >  Download:  30.49 Mbps
> >    Upload:  46.52 Mbps
> >   Latency: (in msec, 61 pings, 0.00% packet loss)
> >       Min: 19.848
> >     10pct: 19.958
> >    Median: 20.743
> >       Avg: 21.229
> >     90pct: 22.739
> >       Max: 30.491
> > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> >  Download:  7.02 Mbps
> >    Upload:  63.68 Mbps
> >   Latency: (in msec, 61 pings, 0.00% packet loss)
> >       Min: 21.690
> >     10pct: 117.997
> >    Median: 153.983
> >       Avg: 150.851
> >     90pct: 180.777
> >       Max: 190.265
> >
> >
> > The first result i gotten by only shaping the uplink and the second results are for only shaping the downlink...
> >
> > and for the testing i put 0 on the for the downlink and uplink when I was testing for the opposite. and I put 50000 in it when I was shaping it for the test. Am I doing it right ?
> >
> > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com> wrote:
> > shape the download only, then measure. shape the upload only, then measure.
> >
> > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> > > Oh OK I mean I was just wondering because I was trying to figure out what
> > > should I have put in the sqm tab because I saw that the wndr3800 tops out at
> > > 50 mbit combined of download and upload right? As of now I have it at 25/25
> > > on the page. Is that right or should I just shape the upload?
> > >
> > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> > >>
> > >> On at least one verizon device I've tried it appeared that they had
> > >> SFQ or something similar on egress from the modem.
> > >>
> > >>
> > >> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > >>
> > >> So you only needed to shape the download. which is good as we start
> > >> peaking out at 50Mbit download total. But only measurements can tell.
> > >>
> > >>
> > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
> > >> wrote:
> > >> > Hello everyone!
> > >> > I have a question about the wndr3800 routing limit. I went back to the
> > >> > older
> > >> > submissions to see if I can find what would be the answer for it. But in
> > >> > my
> > >> > search I haven't managed to find a definite answer. From what I seen
> > >> > about
> > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> > >> > wondering if anyone can shed some light for me here as I have verizon
> > >> > fios
> > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > >> >
> > >> > _______________________________________________
> > >> > Cerowrt-devel mailing list
> > >> > Cerowrt-devel@lists.bufferbloat.net
> > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > >> >
> > >>
> > >>
> > >>
> > >> --
> > >> Dave Täht
> > >>
> > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> >
> >
> >
> > --
> > Dave Täht
> >
> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> >
> > _______________________________________________
> > Cerowrt-devel mailing list
> > Cerowrt-devel@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> 
> 


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-19 22:44       ` Sebastian Moeller
@ 2014-10-20 20:04         ` Dave Taht
  2014-10-20 22:48           ` Dave Taht
  0 siblings, 1 reply; 24+ messages in thread
From: Dave Taht @ 2014-10-20 20:04 UTC (permalink / raw)
  To: Sebastian Moeller; +Cc: cerowrt-devel

On Sun, Oct 19, 2014 at 3:44 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
> Hi Dave,
>
> so I just went with what I have available, shaping between linux host on se00 and macbook on sw10, shaper on se00 (25M/25M that is the maximum wireless throughput with the current router position): burst/cburst set at 1600(default) 16000 and 16000. And lo and behold the sirq gets smaller the larger burst/cburst is set. Now tho test is just too confounded by my bad wireless to be proof, but it certainly justifies the time to expose knobs in the GUI to set burst/cburst/quantum values for HTB for each shaper instance independent for ingress and egress… (I do not assume that this will even double the throughput of a wndr as a router, but even just 10-20% will make a difference ;) (I will eat my own dogwood, since I am about to upgrade from 16M/2.5M to 50M/10M right into where our sirq pain starts)).

I don't necessarily think knobs need to be exposed. Perhaps tuning the
burst parameter as a function of the
induced latency would be about right. At 10mbits, a single 1500 byte
packet takes 1.3ms to egress. So if we
were to aim for .5-2ms worth of burst across the operational range of
the shaper, that might work. In the
case of cable, a grant request takes 2-6ms, anyway.

So at 80mbit, a burst size of 8-16k seems possibly optimal. It could
be higher (other overheads in the kernel).

Now that we have a knob to jiggle, I'll go jiggle it when I have some time...

I note that I'm under the impression cburst can be twiddled with also
to make "powerboost"'s behavior better,
but I've not seen it work.
>
>
>
> Best Regards
>         Sebastian
>
> On Oct 19, 2014, at 21:27 , Dave Taht <dave.taht@gmail.com> wrote:
>
>> Yes fiddling with burst seems to make sense. Try 16k
>>
>> On Oct 19, 2014 11:56 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
>> HI Dave,
>>
>>
>> On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:
>>
>> > On at least one verizon device I've tried it appeared that they had
>> > SFQ or something similar on egress from the modem.
>> >
>> > http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
>> >
>> > So you only needed to shape the download. which is good as we start
>> > peaking out at 50Mbit download total. But only measurements can tell.
>>
>>         So on Hnymans community openwrt build a few fortunate ones on excellent lines seem to get decent results even at 110-120 Mbps combined:
>> https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
>> and:
>> https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
>> I have no idea why and both lines were reasonably well-behaved even without any AQM/QOS...
>>
>> Also I wonder whether when we increase the quantum for higher rates to give HTB some breathing room, whether we also should increase burst and cburst? My hunch is that quantum affects the switching between the leaves, while busts and cburst should allow to dump more data to lower layers inside each leaf qdisc. And since we are running behind, maybe taking a bigger shovel can help some. (I assume this needs to be titrated not to kill latency under load, but if we can only effective have HTB execute x times per second we can easily afford to dump line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet link is way to slow to test this...
>>
>> Best Regards
>>         Sebastian
>>
>> >
>> >
>> > On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
>> >> Hello everyone!
>> >> I have a question about the wndr3800 routing limit. I went back to the older
>> >> submissions to see if I can find what would be the answer for it. But in my
>> >> search I haven't managed to find a definite answer. From what I seen about
>> >> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
>> >> wondering if anyone can shed some light for me here as I have verizon fios
>> >> and my speeds are 50 dl/50 ul. Thank you guys very much!
>> >>
>> >> _______________________________________________
>> >> Cerowrt-devel mailing list
>> >> Cerowrt-devel@lists.bufferbloat.net
>> >> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>> >>
>> >
>> >
>> >
>> > --
>> > Dave Täht
>> >
>> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>> > _______________________________________________
>> > Cerowrt-devel mailing list
>> > Cerowrt-devel@lists.bufferbloat.net
>> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>
>



-- 
Dave Täht

thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-20 20:04         ` Dave Taht
@ 2014-10-20 22:48           ` Dave Taht
  2014-10-21  3:35             ` Aaron Wood
  2014-10-21  7:51             ` Sebastian Moeller
  0 siblings, 2 replies; 24+ messages in thread
From: Dave Taht @ 2014-10-20 22:48 UTC (permalink / raw)
  To: Sebastian Moeller; +Cc: cerowrt-devel

OK, I tried a few combinations of burst and cburst on a cerowrt box,
using 90/10 as a up/download speed.

burst 64000 cburst 64000 was a bit of a win, in most respects, but odd
in others.

netperf-wrapper data at:

http://snapon.lab.bufferbloat.net/~d/burst_tests/


On Mon, Oct 20, 2014 at 1:04 PM, Dave Taht <dave.taht@gmail.com> wrote:
> On Sun, Oct 19, 2014 at 3:44 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
>> Hi Dave,
>>
>> so I just went with what I have available, shaping between linux host on se00 and macbook on sw10, shaper on se00 (25M/25M that is the maximum wireless throughput with the current router position): burst/cburst set at 1600(default) 16000 and 16000. And lo and behold the sirq gets smaller the larger burst/cburst is set. Now tho test is just too confounded by my bad wireless to be proof, but it certainly justifies the time to expose knobs in the GUI to set burst/cburst/quantum values for HTB for each shaper instance independent for ingress and egress… (I do not assume that this will even double the throughput of a wndr as a router, but even just 10-20% will make a difference ;) (I will eat my own dogwood, since I am about to upgrade from 16M/2.5M to 50M/10M right into where our sirq pain starts)).
>
> I don't necessarily think knobs need to be exposed. Perhaps tuning the
> burst parameter as a function of the
> induced latency would be about right. At 10mbits, a single 1500 byte
> packet takes 1.3ms to egress. So if we
> were to aim for .5-2ms worth of burst across the operational range of
> the shaper, that might work. In the
> case of cable, a grant request takes 2-6ms, anyway.
>
> So at 80mbit, a burst size of 8-16k seems possibly optimal. It could
> be higher (other overheads in the kernel).
>
> Now that we have a knob to jiggle, I'll go jiggle it when I have some time...
>
> I note that I'm under the impression cburst can be twiddled with also
> to make "powerboost"'s behavior better,
> but I've not seen it work.
>>
>>
>>
>> Best Regards
>>         Sebastian
>>
>> On Oct 19, 2014, at 21:27 , Dave Taht <dave.taht@gmail.com> wrote:
>>
>>> Yes fiddling with burst seems to make sense. Try 16k
>>>
>>> On Oct 19, 2014 11:56 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
>>> HI Dave,
>>>
>>>
>>> On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:
>>>
>>> > On at least one verizon device I've tried it appeared that they had
>>> > SFQ or something similar on egress from the modem.
>>> >
>>> > http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
>>> >
>>> > So you only needed to shape the download. which is good as we start
>>> > peaking out at 50Mbit download total. But only measurements can tell.
>>>
>>>         So on Hnymans community openwrt build a few fortunate ones on excellent lines seem to get decent results even at 110-120 Mbps combined:
>>> https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
>>> and:
>>> https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
>>> I have no idea why and both lines were reasonably well-behaved even without any AQM/QOS...
>>>
>>> Also I wonder whether when we increase the quantum for higher rates to give HTB some breathing room, whether we also should increase burst and cburst? My hunch is that quantum affects the switching between the leaves, while busts and cburst should allow to dump more data to lower layers inside each leaf qdisc. And since we are running behind, maybe taking a bigger shovel can help some. (I assume this needs to be titrated not to kill latency under load, but if we can only effective have HTB execute x times per second we can easily afford to dump line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet link is way to slow to test this...
>>>
>>> Best Regards
>>>         Sebastian
>>>
>>> >
>>> >
>>> > On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
>>> >> Hello everyone!
>>> >> I have a question about the wndr3800 routing limit. I went back to the older
>>> >> submissions to see if I can find what would be the answer for it. But in my
>>> >> search I haven't managed to find a definite answer. From what I seen about
>>> >> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
>>> >> wondering if anyone can shed some light for me here as I have verizon fios
>>> >> and my speeds are 50 dl/50 ul. Thank you guys very much!
>>> >>
>>> >> _______________________________________________
>>> >> Cerowrt-devel mailing list
>>> >> Cerowrt-devel@lists.bufferbloat.net
>>> >> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>> >>
>>> >
>>> >
>>> >
>>> > --
>>> > Dave Täht
>>> >
>>> > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>>> > _______________________________________________
>>> > Cerowrt-devel mailing list
>>> > Cerowrt-devel@lists.bufferbloat.net
>>> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>>
>>
>
>
>
> --
> Dave Täht
>
> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks



-- 
Dave Täht

thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-20  8:11             ` Sebastian Moeller
@ 2014-10-21  2:30               ` Ernesto Elias
  2014-10-21  8:23                 ` Sebastian Moeller
  0 siblings, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-21  2:30 UTC (permalink / raw)
  To: Sebastian Moeller; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 6812 bytes --]

Alright Sebastian, I've managed to get the virtual box and Ubuntu running
on the computer and I went to go see the packages I have no idea which one
to install there's a whole lot of them. Forgive me this is all new to me.
Which one do I acquire and how to install it?
On Oct 20, 2014 4:11 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:

> Hi Ernesto,
>
> On Oct 20, 2014, at 03:08 , Ernesto Elias <ernestogelias@gmail.com> wrote:
>
> >
> > Well Sebastian sir, I am taking your advice and I'm on windows 7 so
> right now I'm downloading virtualbox and xubuntu. I don't know if there is
> one that you guys recommend.
>
>         I only tested ubuntu 12.04LTS, but certainly I would try the
> current 14.07 LTS today.
>
> >  Well... I am new to all this stuff so I'm asking you to please help me
> on the subject like after everything is up and running on the computer,
>
>         Oh, sure I am happy to help you with getting better data ;)
>
> > what other steps do I do afterwards and forgive me like I said I'm new
> ot all this stuff because I want to be able to find my link capacity when
> using SQM. and Thanks for the help!
>
>         First step is getting the vitalized linux up and running, then
> python 2.7 (I assume that python will be part of the distribution), with
> matplotlib; and finally netperf-wrapper. I assume it might take a bit
> trying to figure out all the required packets installed, but it should be
> reasonably quick to get this up and running… Just give it a try and let me
> know if/where you get stuck
>         It would be most excellent if you could document the whole
> procedure somehow so we could put it somewhere on the net to make it easier
> in the future...
>
> Best Regards
>         Sebastian
>
> >
> > On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de>
> wrote:
> > Hi Ernesto,
> >
> > I would recommend to not run the netperf processes on the router you
> want to assess, if possible. Shaping at the rates you are interested in
> probably will take 100% CPU of the router, so there is no slack for running
> net server and netperf instances at the same time. Also I would recommend
> to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper)
> instead of netperfrunner.sh as it allows so much nicer visualization of the
> queueing over time. netperfrunner.sh is great tool for quick testing and
> figuring out link capacities and worst case latency under load increase
> (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper
> will run under linux and macosx natively, but will also run (well enough)
> in a linux virtual machine on a windows box.
> >
> >
> > Best Regards
> >         Sebastian
> >
> >
> > On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> >
> > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> > >  Download:  30.49 Mbps
> > >    Upload:  46.52 Mbps
> > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > >       Min: 19.848
> > >     10pct: 19.958
> > >    Median: 20.743
> > >       Avg: 21.229
> > >     90pct: 22.739
> > >       Max: 30.491
> > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> > >  Download:  7.02 Mbps
> > >    Upload:  63.68 Mbps
> > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > >       Min: 21.690
> > >     10pct: 117.997
> > >    Median: 153.983
> > >       Avg: 150.851
> > >     90pct: 180.777
> > >       Max: 190.265
> > >
> > >
> > > The first result i gotten by only shaping the uplink and the second
> results are for only shaping the downlink...
> > >
> > > and for the testing i put 0 on the for the downlink and uplink when I
> was testing for the opposite. and I put 50000 in it when I was shaping it
> for the test. Am I doing it right ?
> > >
> > > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com>
> wrote:
> > > shape the download only, then measure. shape the upload only, then
> measure.
> > >
> > > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <
> ernestogelias@gmail.com> wrote:
> > > > Oh OK I mean I was just wondering because I was trying to figure out
> what
> > > > should I have put in the sqm tab because I saw that the wndr3800
> tops out at
> > > > 50 mbit combined of download and upload right? As of now I have it
> at 25/25
> > > > on the page. Is that right or should I just shape the upload?
> > > >
> > > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> > > >>
> > > >> On at least one verizon device I've tried it appeared that they had
> > > >> SFQ or something similar on egress from the modem.
> > > >>
> > > >>
> > > >>
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > > >>
> > > >> So you only needed to shape the download. which is good as we start
> > > >> peaking out at 50Mbit download total. But only measurements can
> tell.
> > > >>
> > > >>
> > > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <
> ernestogelias@gmail.com>
> > > >> wrote:
> > > >> > Hello everyone!
> > > >> > I have a question about the wndr3800 routing limit. I went back
> to the
> > > >> > older
> > > >> > submissions to see if I can find what would be the answer for it.
> But in
> > > >> > my
> > > >> > search I haven't managed to find a definite answer. From what I
> seen
> > > >> > about
> > > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm
> just
> > > >> > wondering if anyone can shed some light for me here as I have
> verizon
> > > >> > fios
> > > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > > >> >
> > > >> > _______________________________________________
> > > >> > Cerowrt-devel mailing list
> > > >> > Cerowrt-devel@lists.bufferbloat.net
> > > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > >> >
> > > >>
> > > >>
> > > >>
> > > >> --
> > > >> Dave Täht
> > > >>
> > > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > >
> > >
> > >
> > > --
> > > Dave Täht
> > >
> > > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > >
> > > _______________________________________________
> > > Cerowrt-devel mailing list
> > > Cerowrt-devel@lists.bufferbloat.net
> > > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> >
> >
>
>

[-- Attachment #2: Type: text/html, Size: 9826 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-20 22:48           ` Dave Taht
@ 2014-10-21  3:35             ` Aaron Wood
  2014-10-21  7:43               ` Sebastian Moeller
  2014-10-21  7:51             ` Sebastian Moeller
  1 sibling, 1 reply; 24+ messages in thread
From: Aaron Wood @ 2014-10-21  3:35 UTC (permalink / raw)
  To: Dave Taht; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 367 bytes --]

On Mon, Oct 20, 2014 at 3:48 PM, Dave Taht <dave.taht@gmail.com> wrote:

> OK, I tried a few combinations of burst and cburst on a cerowrt box,
> using 90/10 as a up/download speed.
>
> burst 64000 cburst 64000 was a bit of a win, in most respects, but odd
> in others.
>

Interesting....  can this be specified on the advanced options string
(ingress only)?

-Aaron

[-- Attachment #2: Type: text/html, Size: 718 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-21  3:35             ` Aaron Wood
@ 2014-10-21  7:43               ` Sebastian Moeller
  0 siblings, 0 replies; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-21  7:43 UTC (permalink / raw)
  To: Aaron Wood; +Cc: cerowrt-devel

Hi Aaron,


On Oct 21, 2014, at 05:35 , Aaron Wood <woody77@gmail.com> wrote:

> On Mon, Oct 20, 2014 at 3:48 PM, Dave Taht <dave.taht@gmail.com> wrote:
> OK, I tried a few combinations of burst and cburst on a cerowrt box,
> using 90/10 as a up/download speed.
> 
> burst 64000 cburst 64000 was a bit of a win, in most respects, but odd
> in others.
> 
> Interesting....  can this be specified on the advanced options string (ingress only)? 

	Unfortunately not, the advanced option string is just passed to the leaf qdiscs, but if you are willing to edit simple.qos you can change:
LQ="quantum `get_mtu $IFACE $CEIL`”

into:
LQ="quantum `get_mtu $IFACE $CEIL` burst 64000 cburst 64000” for either ingress or egress or both to get the desired effect. As always “tc class show dev ge00” and "tc class show dev ifb4ge00” should show you wether your changes took hold (assuming you use a recent sqm-scripts and ge00…).
	So I probably will try to expose these options in an additional optional hidden-by-default entry fields in a future revision.


Best Regards
	Sebastian

> 
> -Aaron


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-20 22:48           ` Dave Taht
  2014-10-21  3:35             ` Aaron Wood
@ 2014-10-21  7:51             ` Sebastian Moeller
  1 sibling, 0 replies; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-21  7:51 UTC (permalink / raw)
  To: Dave Täht; +Cc: cerowrt-devel

HI Dave,


On Oct 21, 2014, at 00:48 , Dave Taht <dave.taht@gmail.com> wrote:

> OK, I tried a few combinations of burst and cburst on a cerowrt box,
> using 90/10 as a up/download speed.
> 
> burst 64000 cburst 64000 was a bit of a win, in most respects, but odd
> in others.
> 
> netperf-wrapper data at:
> 
> http://snapon.lab.bufferbloat.net/~d/burst_tests/

	Great, so I see an increase in ingress rate for cburst 64000 with no real side effects, but burst 64000 cburst 64000 somehow manages to punish one of the BE streams. Luckily the latency under load increase stays really small. Do you know whether sirq differed between the tests? (I assume no it should be pegged to 90-100% during all tests as the effective good put increased…)

Best Regards
        Sebastian

> 
> 
> On Mon, Oct 20, 2014 at 1:04 PM, Dave Taht <dave.taht@gmail.com> wrote:
>> On Sun, Oct 19, 2014 at 3:44 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
>>> Hi Dave,
>>> 
>>> so I just went with what I have available, shaping between linux host on se00 and macbook on sw10, shaper on se00 (25M/25M that is the maximum wireless throughput with the current router position): burst/cburst set at 1600(default) 16000 and 16000. And lo and behold the sirq gets smaller the larger burst/cburst is set. Now tho test is just too confounded by my bad wireless to be proof, but it certainly justifies the time to expose knobs in the GUI to set burst/cburst/quantum values for HTB for each shaper instance independent for ingress and egress… (I do not assume that this will even double the throughput of a wndr as a router, but even just 10-20% will make a difference ;) (I will eat my own dogwood, since I am about to upgrade from 16M/2.5M to 50M/10M right into where our sirq pain starts)).
>> 
>> I don't necessarily think knobs need to be exposed. Perhaps tuning the
>> burst parameter as a function of the
>> induced latency would be about right. At 10mbits, a single 1500 byte
>> packet takes 1.3ms to egress. So if we
>> were to aim for .5-2ms worth of burst across the operational range of
>> the shaper, that might work. In the
>> case of cable, a grant request takes 2-6ms, anyway.
>> 
>> So at 80mbit, a burst size of 8-16k seems possibly optimal. It could
>> be higher (other overheads in the kernel).
>> 
>> Now that we have a knob to jiggle, I'll go jiggle it when I have some time...
>> 
>> I note that I'm under the impression cburst can be twiddled with also
>> to make "powerboost"'s behavior better,
>> but I've not seen it work.
>>> 
>>> 
>>> 
>>> Best Regards
>>>        Sebastian
>>> 
>>> On Oct 19, 2014, at 21:27 , Dave Taht <dave.taht@gmail.com> wrote:
>>> 
>>>> Yes fiddling with burst seems to make sense. Try 16k
>>>> 
>>>> On Oct 19, 2014 11:56 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
>>>> HI Dave,
>>>> 
>>>> 
>>>> On Oct 19, 2014, at 20:24 , Dave Taht <dave.taht@gmail.com> wrote:
>>>> 
>>>>> On at least one verizon device I've tried it appeared that they had
>>>>> SFQ or something similar on egress from the modem.
>>>>> 
>>>>> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
>>>>> 
>>>>> So you only needed to shape the download. which is good as we start
>>>>> peaking out at 50Mbit download total. But only measurements can tell.
>>>> 
>>>>        So on Hnymans community openwrt build a few fortunate ones on excellent lines seem to get decent results even at 110-120 Mbps combined:
>>>> https://forum.openwrt.org/viewtopic.php?pid=250989#p250989
>>>> and:
>>>> https://forum.openwrt.org/viewtopic.php?pid=251013#p251013
>>>> I have no idea why and both lines were reasonably well-behaved even without any AQM/QOS...
>>>> 
>>>> Also I wonder whether when we increase the quantum for higher rates to give HTB some breathing room, whether we also should increase burst and cburst? My hunch is that quantum affects the switching between the leaves, while busts and cburst should allow to dump more data to lower layers inside each leaf qdisc. And since we are running behind, maybe taking a bigger shovel can help some. (I assume this needs to be titrated not to kill latency under load, but if we can only effective have HTB execute x times per second we can easily afford to dump line-rate/maxHTB_iteratin_rate bytes per opportunity, no?) My own internet link is way to slow to test this...
>>>> 
>>>> Best Regards
>>>>        Sebastian
>>>> 
>>>>> 
>>>>> 
>>>>> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
>>>>>> Hello everyone!
>>>>>> I have a question about the wndr3800 routing limit. I went back to the older
>>>>>> submissions to see if I can find what would be the answer for it. But in my
>>>>>> search I haven't managed to find a definite answer. From what I seen about
>>>>>> setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
>>>>>> wondering if anyone can shed some light for me here as I have verizon fios
>>>>>> and my speeds are 50 dl/50 ul. Thank you guys very much!
>>>>>> 
>>>>>> _______________________________________________
>>>>>> Cerowrt-devel mailing list
>>>>>> Cerowrt-devel@lists.bufferbloat.net
>>>>>> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> --
>>>>> Dave Täht
>>>>> 
>>>>> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
>>>>> _______________________________________________
>>>>> Cerowrt-devel mailing list
>>>>> Cerowrt-devel@lists.bufferbloat.net
>>>>> https://lists.bufferbloat.net/listinfo/cerowrt-devel
>>>> 
>>> 
>> 
>> 
>> 
>> --
>> Dave Täht
>> 
>> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> 
> 
> 
> -- 
> Dave Täht
> 
> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-21  2:30               ` Ernesto Elias
@ 2014-10-21  8:23                 ` Sebastian Moeller
  2014-10-22 19:39                   ` Ernesto Elias
  0 siblings, 1 reply; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-21  8:23 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

Hi Ernesto,

so I would recommend to first install git from a terminal in your ubuntu virtual machine:
	sudo apt-get install git

then create a directory where to store horse netperf-wrapper:
	mkdir -p ~/bufferbloat ; cd ~/bufferbloat

then clone the netperf-wrapper repository:
	git clone git@https://github.com/tohojo/netperf-wrapper

now you can test what is missing by calling:
	./netperf-wrapper
this will most likely complain about missing components, please send the output of this command so we can take it from there.

In the unexpected case that the above command just success please try:
	date ; ping -c 10 netperf-eu.bufferbloat.net ; ./netperf-wrapper --ipv4 -l 300 -H netperf-eu.bufferbloat.net rrul -p all_scaled --disable-log -t netperf-wrapper_test 

and again send the output…

And finally if all the above succeeds please try 
	./netperf-wrapper —gui

and send the output

Hope that helps...

Best Regards
	Sebastian



On Oct 21, 2014, at 04:30 , Ernesto Elias <ernestogelias@gmail.com> wrote:

> Alright Sebastian, I've managed to get the virtual box and Ubuntu running on the computer and I went to go see the packages I have no idea which one to install there's a whole lot of them. Forgive me this is all new to me. Which one do I acquire and how to install it?
> 
> On Oct 20, 2014 4:11 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> Hi Ernesto,
> 
> On Oct 20, 2014, at 03:08 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> 
> >
> > Well Sebastian sir, I am taking your advice and I'm on windows 7 so right now I'm downloading virtualbox and xubuntu. I don't know if there is one that you guys recommend.
> 
>         I only tested ubuntu 12.04LTS, but certainly I would try the current 14.07 LTS today.
> 
> >  Well... I am new to all this stuff so I'm asking you to please help me on the subject like after everything is up and running on the computer,
> 
>         Oh, sure I am happy to help you with getting better data ;)
> 
> > what other steps do I do afterwards and forgive me like I said I'm new ot all this stuff because I want to be able to find my link capacity when using SQM. and Thanks for the help!
> 
>         First step is getting the vitalized linux up and running, then python 2.7 (I assume that python will be part of the distribution), with matplotlib; and finally netperf-wrapper. I assume it might take a bit trying to figure out all the required packets installed, but it should be reasonably quick to get this up and running… Just give it a try and let me know if/where you get stuck
>         It would be most excellent if you could document the whole procedure somehow so we could put it somewhere on the net to make it easier in the future...
> 
> Best Regards
>         Sebastian
> 
> >
> > On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
> > Hi Ernesto,
> >
> > I would recommend to not run the netperf processes on the router you want to assess, if possible. Shaping at the rates you are interested in probably will take 100% CPU of the router, so there is no slack for running net server and netperf instances at the same time. Also I would recommend to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper) instead of netperfrunner.sh as it allows so much nicer visualization of the queueing over time. netperfrunner.sh is great tool for quick testing and figuring out link capacities and worst case latency under load increase (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper will run under linux and macosx natively, but will also run (well enough) in a linux virtual machine on a windows box.
> >
> >
> > Best Regards
> >         Sebastian
> >
> >
> > On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> >
> > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> > >  Download:  30.49 Mbps
> > >    Upload:  46.52 Mbps
> > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > >       Min: 19.848
> > >     10pct: 19.958
> > >    Median: 20.743
> > >       Avg: 21.229
> > >     90pct: 22.739
> > >       Max: 30.491
> > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> > >  Download:  7.02 Mbps
> > >    Upload:  63.68 Mbps
> > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > >       Min: 21.690
> > >     10pct: 117.997
> > >    Median: 153.983
> > >       Avg: 150.851
> > >     90pct: 180.777
> > >       Max: 190.265
> > >
> > >
> > > The first result i gotten by only shaping the uplink and the second results are for only shaping the downlink...
> > >
> > > and for the testing i put 0 on the for the downlink and uplink when I was testing for the opposite. and I put 50000 in it when I was shaping it for the test. Am I doing it right ?
> > >
> > > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com> wrote:
> > > shape the download only, then measure. shape the upload only, then measure.
> > >
> > > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> > > > Oh OK I mean I was just wondering because I was trying to figure out what
> > > > should I have put in the sqm tab because I saw that the wndr3800 tops out at
> > > > 50 mbit combined of download and upload right? As of now I have it at 25/25
> > > > on the page. Is that right or should I just shape the upload?
> > > >
> > > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> > > >>
> > > >> On at least one verizon device I've tried it appeared that they had
> > > >> SFQ or something similar on egress from the modem.
> > > >>
> > > >>
> > > >> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > > >>
> > > >> So you only needed to shape the download. which is good as we start
> > > >> peaking out at 50Mbit download total. But only measurements can tell.
> > > >>
> > > >>
> > > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
> > > >> wrote:
> > > >> > Hello everyone!
> > > >> > I have a question about the wndr3800 routing limit. I went back to the
> > > >> > older
> > > >> > submissions to see if I can find what would be the answer for it. But in
> > > >> > my
> > > >> > search I haven't managed to find a definite answer. From what I seen
> > > >> > about
> > > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> > > >> > wondering if anyone can shed some light for me here as I have verizon
> > > >> > fios
> > > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > > >> >
> > > >> > _______________________________________________
> > > >> > Cerowrt-devel mailing list
> > > >> > Cerowrt-devel@lists.bufferbloat.net
> > > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > >> >
> > > >>
> > > >>
> > > >>
> > > >> --
> > > >> Dave Täht
> > > >>
> > > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > >
> > >
> > >
> > > --
> > > Dave Täht
> > >
> > > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > >
> > > _______________________________________________
> > > Cerowrt-devel mailing list
> > > Cerowrt-devel@lists.bufferbloat.net
> > > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> >
> >
> 


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-21  8:23                 ` Sebastian Moeller
@ 2014-10-22 19:39                   ` Ernesto Elias
  2014-10-22 20:36                     ` Sebastian Moeller
  0 siblings, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-22 19:39 UTC (permalink / raw)
  To: Sebastian Moeller; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 8522 bytes --]

Hello Sebastian, well I followed your directions and when I inputted the
command git clone git at https://github.com/tohojo/netperf-wrapper

It said too many arguments...
What do I do now?
On Oct 21, 2014 4:23 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:

> Hi Ernesto,
>
> so I would recommend to first install git from a terminal in your ubuntu
> virtual machine:
>         sudo apt-get install git
>
> then create a directory where to store horse netperf-wrapper:
>         mkdir -p ~/bufferbloat ; cd ~/bufferbloat
>
> then clone the netperf-wrapper repository:
>         git clone git@https://github.com/tohojo/netperf-wrapper
>
> now you can test what is missing by calling:
>         ./netperf-wrapper
> this will most likely complain about missing components, please send the
> output of this command so we can take it from there.
>
> In the unexpected case that the above command just success please try:
>         date ; ping -c 10 netperf-eu.bufferbloat.net ; ./netperf-wrapper
> --ipv4 -l 300 -H netperf-eu.bufferbloat.net rrul -p all_scaled
> --disable-log -t netperf-wrapper_test
>
> and again send the output…
>
> And finally if all the above succeeds please try
>         ./netperf-wrapper —gui
>
> and send the output
>
> Hope that helps...
>
> Best Regards
>         Sebastian
>
>
>
> On Oct 21, 2014, at 04:30 , Ernesto Elias <ernestogelias@gmail.com> wrote:
>
> > Alright Sebastian, I've managed to get the virtual box and Ubuntu
> running on the computer and I went to go see the packages I have no idea
> which one to install there's a whole lot of them. Forgive me this is all
> new to me. Which one do I acquire and how to install it?
> >
> > On Oct 20, 2014 4:11 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> > Hi Ernesto,
> >
> > On Oct 20, 2014, at 03:08 , Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> >
> > >
> > > Well Sebastian sir, I am taking your advice and I'm on windows 7 so
> right now I'm downloading virtualbox and xubuntu. I don't know if there is
> one that you guys recommend.
> >
> >         I only tested ubuntu 12.04LTS, but certainly I would try the
> current 14.07 LTS today.
> >
> > >  Well... I am new to all this stuff so I'm asking you to please help
> me on the subject like after everything is up and running on the computer,
> >
> >         Oh, sure I am happy to help you with getting better data ;)
> >
> > > what other steps do I do afterwards and forgive me like I said I'm new
> ot all this stuff because I want to be able to find my link capacity when
> using SQM. and Thanks for the help!
> >
> >         First step is getting the vitalized linux up and running, then
> python 2.7 (I assume that python will be part of the distribution), with
> matplotlib; and finally netperf-wrapper. I assume it might take a bit
> trying to figure out all the required packets installed, but it should be
> reasonably quick to get this up and running… Just give it a try and let me
> know if/where you get stuck
> >         It would be most excellent if you could document the whole
> procedure somehow so we could put it somewhere on the net to make it easier
> in the future...
> >
> > Best Regards
> >         Sebastian
> >
> > >
> > > On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de>
> wrote:
> > > Hi Ernesto,
> > >
> > > I would recommend to not run the netperf processes on the router you
> want to assess, if possible. Shaping at the rates you are interested in
> probably will take 100% CPU of the router, so there is no slack for running
> net server and netperf instances at the same time. Also I would recommend
> to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper)
> instead of netperfrunner.sh as it allows so much nicer visualization of the
> queueing over time. netperfrunner.sh is great tool for quick testing and
> figuring out link capacities and worst case latency under load increase
> (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper
> will run under linux and macosx natively, but will also run (well enough)
> in a linux virtual machine on a windows box.
> > >
> > >
> > > Best Regards
> > >         Sebastian
> > >
> > >
> > > On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> > >
> > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> > > >  Download:  30.49 Mbps
> > > >    Upload:  46.52 Mbps
> > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > >       Min: 19.848
> > > >     10pct: 19.958
> > > >    Median: 20.743
> > > >       Avg: 21.229
> > > >     90pct: 22.739
> > > >       Max: 30.491
> > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> > > >  Download:  7.02 Mbps
> > > >    Upload:  63.68 Mbps
> > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > >       Min: 21.690
> > > >     10pct: 117.997
> > > >    Median: 153.983
> > > >       Avg: 150.851
> > > >     90pct: 180.777
> > > >       Max: 190.265
> > > >
> > > >
> > > > The first result i gotten by only shaping the uplink and the second
> results are for only shaping the downlink...
> > > >
> > > > and for the testing i put 0 on the for the downlink and uplink when
> I was testing for the opposite. and I put 50000 in it when I was shaping it
> for the test. Am I doing it right ?
> > > >
> > > > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com>
> wrote:
> > > > shape the download only, then measure. shape the upload only, then
> measure.
> > > >
> > > > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <
> ernestogelias@gmail.com> wrote:
> > > > > Oh OK I mean I was just wondering because I was trying to figure
> out what
> > > > > should I have put in the sqm tab because I saw that the wndr3800
> tops out at
> > > > > 50 mbit combined of download and upload right? As of now I have it
> at 25/25
> > > > > on the page. Is that right or should I just shape the upload?
> > > > >
> > > > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> > > > >>
> > > > >> On at least one verizon device I've tried it appeared that they
> had
> > > > >> SFQ or something similar on egress from the modem.
> > > > >>
> > > > >>
> > > > >>
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > > > >>
> > > > >> So you only needed to shape the download. which is good as we
> start
> > > > >> peaking out at 50Mbit download total. But only measurements can
> tell.
> > > > >>
> > > > >>
> > > > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <
> ernestogelias@gmail.com>
> > > > >> wrote:
> > > > >> > Hello everyone!
> > > > >> > I have a question about the wndr3800 routing limit. I went back
> to the
> > > > >> > older
> > > > >> > submissions to see if I can find what would be the answer for
> it. But in
> > > > >> > my
> > > > >> > search I haven't managed to find a definite answer. From what I
> seen
> > > > >> > about
> > > > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm
> just
> > > > >> > wondering if anyone can shed some light for me here as I have
> verizon
> > > > >> > fios
> > > > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > > > >> >
> > > > >> > _______________________________________________
> > > > >> > Cerowrt-devel mailing list
> > > > >> > Cerowrt-devel@lists.bufferbloat.net
> > > > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > > >> >
> > > > >>
> > > > >>
> > > > >>
> > > > >> --
> > > > >> Dave Täht
> > > > >>
> > > > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > >
> > > >
> > > >
> > > > --
> > > > Dave Täht
> > > >
> > > > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > >
> > > > _______________________________________________
> > > > Cerowrt-devel mailing list
> > > > Cerowrt-devel@lists.bufferbloat.net
> > > > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > >
> > >
> >
>
>

[-- Attachment #2: Type: text/html, Size: 12384 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-22 19:39                   ` Ernesto Elias
@ 2014-10-22 20:36                     ` Sebastian Moeller
  2014-10-23 22:03                       ` Ernesto Elias
  0 siblings, 1 reply; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-22 20:36 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

Hi Ernesto,



On Oct 22, 2014, at 21:39 , Ernesto Elias <ernestogelias@gmail.com> wrote:

> Hello Sebastian, well I followed your directions and when I inputted the command git clone git at https://github.com/tohojo/netperf-wrapper

	Oops, sorry, it looks I copy and pasted the wrong line from my .bash_history… the following should work:

git clone https://github.com/tohojo/netperf-wrapper

also note Dave’s mail where he recommends:
sudo apt-get install fping python-qt4 python-matplotlib

to get more of the requirements for netperf-wrapper installed,  please run this after coning the netperf-wrapper repository.

Thanks a lot…

	Sebastian


> 
> It said too many arguments... 
> What do I do now?
> 
> On Oct 21, 2014 4:23 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> Hi Ernesto,
> 
> so I would recommend to first install git from a terminal in your ubuntu virtual machine:
>         sudo apt-get install git
> 
> then create a directory where to store horse netperf-wrapper:
>         mkdir -p ~/bufferbloat ; cd ~/bufferbloat
> 
> then clone the netperf-wrapper repository:
>         git clone git@https://github.com/tohojo/netperf-wrapper
> 
> now you can test what is missing by calling:
>         ./netperf-wrapper
> this will most likely complain about missing components, please send the output of this command so we can take it from there.
> 
> In the unexpected case that the above command just success please try:
>         date ; ping -c 10 netperf-eu.bufferbloat.net ; ./netperf-wrapper --ipv4 -l 300 -H netperf-eu.bufferbloat.net rrul -p all_scaled --disable-log -t netperf-wrapper_test
> 
> and again send the output…
> 
> And finally if all the above succeeds please try
>         ./netperf-wrapper —gui
> 
> and send the output
> 
> Hope that helps...
> 
> Best Regards
>         Sebastian
> 
> 
> 
> On Oct 21, 2014, at 04:30 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> 
> > Alright Sebastian, I've managed to get the virtual box and Ubuntu running on the computer and I went to go see the packages I have no idea which one to install there's a whole lot of them. Forgive me this is all new to me. Which one do I acquire and how to install it?
> >
> > On Oct 20, 2014 4:11 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> > Hi Ernesto,
> >
> > On Oct 20, 2014, at 03:08 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> >
> > >
> > > Well Sebastian sir, I am taking your advice and I'm on windows 7 so right now I'm downloading virtualbox and xubuntu. I don't know if there is one that you guys recommend.
> >
> >         I only tested ubuntu 12.04LTS, but certainly I would try the current 14.07 LTS today.
> >
> > >  Well... I am new to all this stuff so I'm asking you to please help me on the subject like after everything is up and running on the computer,
> >
> >         Oh, sure I am happy to help you with getting better data ;)
> >
> > > what other steps do I do afterwards and forgive me like I said I'm new ot all this stuff because I want to be able to find my link capacity when using SQM. and Thanks for the help!
> >
> >         First step is getting the vitalized linux up and running, then python 2.7 (I assume that python will be part of the distribution), with matplotlib; and finally netperf-wrapper. I assume it might take a bit trying to figure out all the required packets installed, but it should be reasonably quick to get this up and running… Just give it a try and let me know if/where you get stuck
> >         It would be most excellent if you could document the whole procedure somehow so we could put it somewhere on the net to make it easier in the future...
> >
> > Best Regards
> >         Sebastian
> >
> > >
> > > On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
> > > Hi Ernesto,
> > >
> > > I would recommend to not run the netperf processes on the router you want to assess, if possible. Shaping at the rates you are interested in probably will take 100% CPU of the router, so there is no slack for running net server and netperf instances at the same time. Also I would recommend to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper) instead of netperfrunner.sh as it allows so much nicer visualization of the queueing over time. netperfrunner.sh is great tool for quick testing and figuring out link capacities and worst case latency under load increase (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper will run under linux and macosx natively, but will also run (well enough) in a linux virtual machine on a windows box.
> > >
> > >
> > > Best Regards
> > >         Sebastian
> > >
> > >
> > > On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> > >
> > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> > > >  Download:  30.49 Mbps
> > > >    Upload:  46.52 Mbps
> > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > >       Min: 19.848
> > > >     10pct: 19.958
> > > >    Median: 20.743
> > > >       Avg: 21.229
> > > >     90pct: 22.739
> > > >       Max: 30.491
> > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> > > >  Download:  7.02 Mbps
> > > >    Upload:  63.68 Mbps
> > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > >       Min: 21.690
> > > >     10pct: 117.997
> > > >    Median: 153.983
> > > >       Avg: 150.851
> > > >     90pct: 180.777
> > > >       Max: 190.265
> > > >
> > > >
> > > > The first result i gotten by only shaping the uplink and the second results are for only shaping the downlink...
> > > >
> > > > and for the testing i put 0 on the for the downlink and uplink when I was testing for the opposite. and I put 50000 in it when I was shaping it for the test. Am I doing it right ?
> > > >
> > > > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com> wrote:
> > > > shape the download only, then measure. shape the upload only, then measure.
> > > >
> > > > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> > > > > Oh OK I mean I was just wondering because I was trying to figure out what
> > > > > should I have put in the sqm tab because I saw that the wndr3800 tops out at
> > > > > 50 mbit combined of download and upload right? As of now I have it at 25/25
> > > > > on the page. Is that right or should I just shape the upload?
> > > > >
> > > > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> > > > >>
> > > > >> On at least one verizon device I've tried it appeared that they had
> > > > >> SFQ or something similar on egress from the modem.
> > > > >>
> > > > >>
> > > > >> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > > > >>
> > > > >> So you only needed to shape the download. which is good as we start
> > > > >> peaking out at 50Mbit download total. But only measurements can tell.
> > > > >>
> > > > >>
> > > > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
> > > > >> wrote:
> > > > >> > Hello everyone!
> > > > >> > I have a question about the wndr3800 routing limit. I went back to the
> > > > >> > older
> > > > >> > submissions to see if I can find what would be the answer for it. But in
> > > > >> > my
> > > > >> > search I haven't managed to find a definite answer. From what I seen
> > > > >> > about
> > > > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> > > > >> > wondering if anyone can shed some light for me here as I have verizon
> > > > >> > fios
> > > > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > > > >> >
> > > > >> > _______________________________________________
> > > > >> > Cerowrt-devel mailing list
> > > > >> > Cerowrt-devel@lists.bufferbloat.net
> > > > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > > >> >
> > > > >>
> > > > >>
> > > > >>
> > > > >> --
> > > > >> Dave Täht
> > > > >>
> > > > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > >
> > > >
> > > >
> > > > --
> > > > Dave Täht
> > > >
> > > > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > >
> > > > _______________________________________________
> > > > Cerowrt-devel mailing list
> > > > Cerowrt-devel@lists.bufferbloat.net
> > > > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > >
> > >
> >
> 


^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-22 20:36                     ` Sebastian Moeller
@ 2014-10-23 22:03                       ` Ernesto Elias
  2014-10-24 11:38                         ` Sebastian Moeller
  0 siblings, 1 reply; 24+ messages in thread
From: Ernesto Elias @ 2014-10-23 22:03 UTC (permalink / raw)
  To: Sebastian Moeller; +Cc: cerowrt-devel

[-- Attachment #1: Type: text/plain, Size: 12688 bytes --]

Hello again Sebastian,

this is the output that i have gotten when entering the command and
following your directions

ubuntu@ubuntu:~/bufferbloat$ date ; ping -c 10 netperf-eu.bufferbloat.net ;
./netperf-wrapper --ipv4 -L 300 -H netperf-eu.bufferbloat.net rrul -p
all_scaled --disable-Log -t netperf-wrapper_test
Thu Oct 23 21:55:09 UTC 2014
PING kau.toke.dk (130.243.26.64) 56(84) bytes of data.
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=1 ttl=249
time=106 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=2 ttl=249
time=107 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=3 ttl=249
time=106 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=4 ttl=249
time=106 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=5 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=6 ttl=249
time=106 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=7 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=8 ttl=249
time=106 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=9 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=10 ttl=249
time=112 ms

--- kau.toke.dk ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 9015ms
rtt min/avg/max/mdev = 105.529/106.794/112.146/1.870 ms
bash: ./netperf-wrapper: Is a directory
ubuntu@ubuntu:~/bufferbloat$ date ; ping -c 10 netperf-eu.bufferbloat.net ;
./netperf-wrapper --ipv4 -L 300 -H netperf-eu.bufferbloat.net rrul -p
all_scaled --disable-Log -t netperf-wrapper_test
Thu Oct 23 21:57:19 UTC 2014
PING kau.toke.dk (130.243.26.64) 56(84) bytes of data.
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=1 ttl=249
time=106 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=2 ttl=249
time=106 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=3 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=4 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=5 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=6 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=7 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=8 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=9 ttl=249
time=105 ms
64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=10 ttl=249
time=105 ms

--- kau.toke.dk ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 9015ms
rtt min/avg/max/mdev = 105.008/105.570/106.151/0.559 ms
bash: ./netperf-wrapper: Is a directory



and this is the output of what you told me to put of what Dave recommended

ubuntu@ubuntu:~/bufferbloat$ sudo apt-get install fping python-qt4
python-matplotlib
Reading package lists... Done
Building dependency tree
Reading state information... Done
E: Unable to locate package fping
E: Unable to locate package python-matplotlib

On Wed, Oct 22, 2014 at 8:36 PM, Sebastian Moeller <moeller0@gmx.de> wrote:

> Hi Ernesto,
>
>
>
> On Oct 22, 2014, at 21:39 , Ernesto Elias <ernestogelias@gmail.com> wrote:
>
> > Hello Sebastian, well I followed your directions and when I inputted the
> command git clone git at https://github.com/tohojo/netperf-wrapper
>
>         Oops, sorry, it looks I copy and pasted the wrong line from my
> .bash_history… the following should work:
>
> git clone https://github.com/tohojo/netperf-wrapper
>
> also note Dave’s mail where he recommends:
> sudo apt-get install fping python-qt4 python-matplotlib
>
> to get more of the requirements for netperf-wrapper installed,  please run
> this after coning the netperf-wrapper repository.
>
> Thanks a lot…
>
>         Sebastian
>
>
> >
> > It said too many arguments...
> > What do I do now?
> >
> > On Oct 21, 2014 4:23 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> > Hi Ernesto,
> >
> > so I would recommend to first install git from a terminal in your ubuntu
> virtual machine:
> >         sudo apt-get install git
> >
> > then create a directory where to store horse netperf-wrapper:
> >         mkdir -p ~/bufferbloat ; cd ~/bufferbloat
> >
> > then clone the netperf-wrapper repository:
> >         git clone git@https://github.com/tohojo/netperf-wrapper
> >
> > now you can test what is missing by calling:
> >         ./netperf-wrapper
> > this will most likely complain about missing components, please send the
> output of this command so we can take it from there.
> >
> > In the unexpected case that the above command just success please try:
> >         date ; ping -c 10 netperf-eu.bufferbloat.net ;
> ./netperf-wrapper --ipv4 -l 300 -H netperf-eu.bufferbloat.net rrul -p
> all_scaled --disable-log -t netperf-wrapper_test
> >
> > and again send the output…
> >
> > And finally if all the above succeeds please try
> >         ./netperf-wrapper —gui
> >
> > and send the output
> >
> > Hope that helps...
> >
> > Best Regards
> >         Sebastian
> >
> >
> >
> > On Oct 21, 2014, at 04:30 , Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> >
> > > Alright Sebastian, I've managed to get the virtual box and Ubuntu
> running on the computer and I went to go see the packages I have no idea
> which one to install there's a whole lot of them. Forgive me this is all
> new to me. Which one do I acquire and how to install it?
> > >
> > > On Oct 20, 2014 4:11 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> > > Hi Ernesto,
> > >
> > > On Oct 20, 2014, at 03:08 , Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> > >
> > > >
> > > > Well Sebastian sir, I am taking your advice and I'm on windows 7 so
> right now I'm downloading virtualbox and xubuntu. I don't know if there is
> one that you guys recommend.
> > >
> > >         I only tested ubuntu 12.04LTS, but certainly I would try the
> current 14.07 LTS today.
> > >
> > > >  Well... I am new to all this stuff so I'm asking you to please help
> me on the subject like after everything is up and running on the computer,
> > >
> > >         Oh, sure I am happy to help you with getting better data ;)
> > >
> > > > what other steps do I do afterwards and forgive me like I said I'm
> new ot all this stuff because I want to be able to find my link capacity
> when using SQM. and Thanks for the help!
> > >
> > >         First step is getting the vitalized linux up and running, then
> python 2.7 (I assume that python will be part of the distribution), with
> matplotlib; and finally netperf-wrapper. I assume it might take a bit
> trying to figure out all the required packets installed, but it should be
> reasonably quick to get this up and running… Just give it a try and let me
> know if/where you get stuck
> > >         It would be most excellent if you could document the whole
> procedure somehow so we could put it somewhere on the net to make it easier
> in the future...
> > >
> > > Best Regards
> > >         Sebastian
> > >
> > > >
> > > > On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de>
> wrote:
> > > > Hi Ernesto,
> > > >
> > > > I would recommend to not run the netperf processes on the router you
> want to assess, if possible. Shaping at the rates you are interested in
> probably will take 100% CPU of the router, so there is no slack for running
> net server and netperf instances at the same time. Also I would recommend
> to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper)
> instead of netperfrunner.sh as it allows so much nicer visualization of the
> queueing over time. netperfrunner.sh is great tool for quick testing and
> figuring out link capacities and worst case latency under load increase
> (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper
> will run under linux and macosx natively, but will also run (well enough)
> in a linux virtual machine on a windows box.
> > > >
> > > >
> > > > Best Regards
> > > >         Sebastian
> > > >
> > > >
> > > > On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com>
> wrote:
> > > >
> > > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> > > > >  Download:  30.49 Mbps
> > > > >    Upload:  46.52 Mbps
> > > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > > >       Min: 19.848
> > > > >     10pct: 19.958
> > > > >    Median: 20.743
> > > > >       Avg: 21.229
> > > > >     90pct: 22.739
> > > > >       Max: 30.491
> > > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4
> streams down a
>
>                    nd up while pinging gstatic.com. Takes about 60
> seconds.
> > > > >  Download:  7.02 Mbps
> > > > >    Upload:  63.68 Mbps
> > > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > > >       Min: 21.690
> > > > >     10pct: 117.997
> > > > >    Median: 153.983
> > > > >       Avg: 150.851
> > > > >     90pct: 180.777
> > > > >       Max: 190.265
> > > > >
> > > > >
> > > > > The first result i gotten by only shaping the uplink and the
> second results are for only shaping the downlink...
> > > > >
> > > > > and for the testing i put 0 on the for the downlink and uplink
> when I was testing for the opposite. and I put 50000 in it when I was
> shaping it for the test. Am I doing it right ?
> > > > >
> > > > > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com>
> wrote:
> > > > > shape the download only, then measure. shape the upload only, then
> measure.
> > > > >
> > > > > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <
> ernestogelias@gmail.com> wrote:
> > > > > > Oh OK I mean I was just wondering because I was trying to figure
> out what
> > > > > > should I have put in the sqm tab because I saw that the wndr3800
> tops out at
> > > > > > 50 mbit combined of download and upload right? As of now I have
> it at 25/25
> > > > > > on the page. Is that right or should I just shape the upload?
> > > > > >
> > > > > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com>
> wrote:
> > > > > >>
> > > > > >> On at least one verizon device I've tried it appeared that they
> had
> > > > > >> SFQ or something similar on egress from the modem.
> > > > > >>
> > > > > >>
> > > > > >>
> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > > > > >>
> > > > > >> So you only needed to shape the download. which is good as we
> start
> > > > > >> peaking out at 50Mbit download total. But only measurements can
> tell.
> > > > > >>
> > > > > >>
> > > > > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <
> ernestogelias@gmail.com>
> > > > > >> wrote:
> > > > > >> > Hello everyone!
> > > > > >> > I have a question about the wndr3800 routing limit. I went
> back to the
> > > > > >> > older
> > > > > >> > submissions to see if I can find what would be the answer for
> it. But in
> > > > > >> > my
> > > > > >> > search I haven't managed to find a definite answer. From what
> I seen
> > > > > >> > about
> > > > > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit.
> I'm just
> > > > > >> > wondering if anyone can shed some light for me here as I have
> verizon
> > > > > >> > fios
> > > > > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > > > > >> >
> > > > > >> > _______________________________________________
> > > > > >> > Cerowrt-devel mailing list
> > > > > >> > Cerowrt-devel@lists.bufferbloat.net
> > > > > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > > > >> >
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >> --
> > > > > >> Dave Täht
> > > > > >>
> > > > > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > > >
> > > > >
> > > > >
> > > > > --
> > > > > Dave Täht
> > > > >
> > > > > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > > >
> > > > > _______________________________________________
> > > > > Cerowrt-devel mailing list
> > > > > Cerowrt-devel@lists.bufferbloat.net
> > > > > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > >
> > > >
> > >
> >
>
>

[-- Attachment #2: Type: text/html, Size: 18734 bytes --]

^ permalink raw reply	[flat|nested] 24+ messages in thread

* Re: [Cerowrt-devel] Routing limit question
  2014-10-23 22:03                       ` Ernesto Elias
@ 2014-10-24 11:38                         ` Sebastian Moeller
  0 siblings, 0 replies; 24+ messages in thread
From: Sebastian Moeller @ 2014-10-24 11:38 UTC (permalink / raw)
  To: Ernesto Elias; +Cc: cerowrt-devel

Hi Ernesto,

it seems that I forgot to mention that you need to change into the buffer bloat/netperf-wrapper directory first:

so please try:
cd ~/bufferbloat/netperf-wrapper
followed by:
date ; ping -c 10 netperf-eu.bufferbloat.net ; ./netperf-wrapper --ipv4 -L 300 -H netperf-eu.bufferbloat.net rrul -p all_scaled --disable-Log -t netperf-wrapper_test

that hopefully should start netperf-wraper for good and display an image showing downlink bandwidth, uplink bandwidth and latency over time. Please note that this plot shows the average rates of 4 streams (for up and down), so for your total bandwidth you need to multiply the numbers by 4. 
	Akso note that with “netperf-wrapper —gui” you can select different plot types that can show total bandwidth or box whisker plots (just play with the GUI).


Best Regards
	Sebastian


On Oct 24, 2014, at 00:03 , Ernesto Elias <ernestogelias@gmail.com> wrote:

> Hello again Sebastian,
> 
> this is the output that i have gotten when entering the command and following your directions 
> 
> ubuntu@ubuntu:~/bufferbloat$ date ; ping -c 10 netperf-eu.bufferbloat.net ; ./netperf-wrapper --ipv4 -L 300 -H netperf-eu.bufferbloat.net rrul -p all_scaled --disable-Log -t netperf-wrapper_test
> Thu Oct 23 21:55:09 UTC 2014
> PING kau.toke.dk (130.243.26.64) 56(84) bytes of data.
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=1 ttl=249 time=106 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=2 ttl=249 time=107 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=3 ttl=249 time=106 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=4 ttl=249 time=106 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=5 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=6 ttl=249 time=106 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=7 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=8 ttl=249 time=106 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=9 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=10 ttl=249 time=112 ms
> 
> --- kau.toke.dk ping statistics ---
> 10 packets transmitted, 10 received, 0% packet loss, time 9015ms
> rtt min/avg/max/mdev = 105.529/106.794/112.146/1.870 ms
> bash: ./netperf-wrapper: Is a directory
> ubuntu@ubuntu:~/bufferbloat$ date ; ping -c 10 netperf-eu.bufferbloat.net ; ./netperf-wrapper --ipv4 -L 300 -H netperf-eu.bufferbloat.net rrul -p all_scaled --disable-Log -t netperf-wrapper_test
> Thu Oct 23 21:57:19 UTC 2014
> PING kau.toke.dk (130.243.26.64) 56(84) bytes of data.
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=1 ttl=249 time=106 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=2 ttl=249 time=106 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=3 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=4 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=5 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=6 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=7 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=8 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=9 ttl=249 time=105 ms
> 64 bytes from tohojo-pc.eki.kau.se (130.243.26.64): icmp_seq=10 ttl=249 time=105 ms
> 
> --- kau.toke.dk ping statistics ---
> 10 packets transmitted, 10 received, 0% packet loss, time 9015ms
> rtt min/avg/max/mdev = 105.008/105.570/106.151/0.559 ms
> bash: ./netperf-wrapper: Is a directory
> 
> 
> 
> and this is the output of what you told me to put of what Dave recommended
> 
> ubuntu@ubuntu:~/bufferbloat$ sudo apt-get install fping python-qt4 python-matplotlib
> Reading package lists... Done
> Building dependency tree       
> Reading state information... Done
> E: Unable to locate package fping
> E: Unable to locate package python-matplotlib
> 
> On Wed, Oct 22, 2014 at 8:36 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
> Hi Ernesto,
> 
> 
> 
> On Oct 22, 2014, at 21:39 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> 
> > Hello Sebastian, well I followed your directions and when I inputted the command git clone git at https://github.com/tohojo/netperf-wrapper
> 
>         Oops, sorry, it looks I copy and pasted the wrong line from my .bash_history… the following should work:
> 
> git clone https://github.com/tohojo/netperf-wrapper
> 
> also note Dave’s mail where he recommends:
> sudo apt-get install fping python-qt4 python-matplotlib
> 
> to get more of the requirements for netperf-wrapper installed,  please run this after coning the netperf-wrapper repository.
> 
> Thanks a lot…
> 
>         Sebastian
> 
> 
> >
> > It said too many arguments...
> > What do I do now?
> >
> > On Oct 21, 2014 4:23 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> > Hi Ernesto,
> >
> > so I would recommend to first install git from a terminal in your ubuntu virtual machine:
> >         sudo apt-get install git
> >
> > then create a directory where to store horse netperf-wrapper:
> >         mkdir -p ~/bufferbloat ; cd ~/bufferbloat
> >
> > then clone the netperf-wrapper repository:
> >         git clone git@https://github.com/tohojo/netperf-wrapper
> >
> > now you can test what is missing by calling:
> >         ./netperf-wrapper
> > this will most likely complain about missing components, please send the output of this command so we can take it from there.
> >
> > In the unexpected case that the above command just success please try:
> >         date ; ping -c 10 netperf-eu.bufferbloat.net ; ./netperf-wrapper --ipv4 -l 300 -H netperf-eu.bufferbloat.net rrul -p all_scaled --disable-log -t netperf-wrapper_test
> >
> > and again send the output…
> >
> > And finally if all the above succeeds please try
> >         ./netperf-wrapper —gui
> >
> > and send the output
> >
> > Hope that helps...
> >
> > Best Regards
> >         Sebastian
> >
> >
> >
> > On Oct 21, 2014, at 04:30 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> >
> > > Alright Sebastian, I've managed to get the virtual box and Ubuntu running on the computer and I went to go see the packages I have no idea which one to install there's a whole lot of them. Forgive me this is all new to me. Which one do I acquire and how to install it?
> > >
> > > On Oct 20, 2014 4:11 AM, "Sebastian Moeller" <moeller0@gmx.de> wrote:
> > > Hi Ernesto,
> > >
> > > On Oct 20, 2014, at 03:08 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> > >
> > > >
> > > > Well Sebastian sir, I am taking your advice and I'm on windows 7 so right now I'm downloading virtualbox and xubuntu. I don't know if there is one that you guys recommend.
> > >
> > >         I only tested ubuntu 12.04LTS, but certainly I would try the current 14.07 LTS today.
> > >
> > > >  Well... I am new to all this stuff so I'm asking you to please help me on the subject like after everything is up and running on the computer,
> > >
> > >         Oh, sure I am happy to help you with getting better data ;)
> > >
> > > > what other steps do I do afterwards and forgive me like I said I'm new ot all this stuff because I want to be able to find my link capacity when using SQM. and Thanks for the help!
> > >
> > >         First step is getting the vitalized linux up and running, then python 2.7 (I assume that python will be part of the distribution), with matplotlib; and finally netperf-wrapper. I assume it might take a bit trying to figure out all the required packets installed, but it should be reasonably quick to get this up and running… Just give it a try and let me know if/where you get stuck
> > >         It would be most excellent if you could document the whole procedure somehow so we could put it somewhere on the net to make it easier in the future...
> > >
> > > Best Regards
> > >         Sebastian
> > >
> > > >
> > > > On Sun, Oct 19, 2014 at 3:55 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
> > > > Hi Ernesto,
> > > >
> > > > I would recommend to not run the netperf processes on the router you want to assess, if possible. Shaping at the rates you are interested in probably will take 100% CPU of the router, so there is no slack for running net server and netperf instances at the same time. Also I would recommend to use netperf-wrapper (https://github.com/tohojo/netperf-wrapper) instead of netperfrunner.sh as it allows so much nicer visualization of the queueing over time. netperfrunner.sh is great tool for quick testing and figuring out link capacities and worst case latency under load increase (LULI), but for more thorough comparisons it ain’t made. Netperf-wrapper will run under linux and macosx natively, but will also run (well enough) in a linux virtual machine on a windows box.
> > > >
> > > >
> > > > Best Regards
> > > >         Sebastian
> > > >
> > > >
> > > > On Oct 19, 2014, at 21:16 , Ernesto Elias <ernestogelias@gmail.com> wrote:
> > > >
> > > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > > 2014-10-19 15:10:27 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> > > > >  Download:  30.49 Mbps
> > > > >    Upload:  46.52 Mbps
> > > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > > >       Min: 19.848
> > > > >     10pct: 19.958
> > > > >    Median: 20.743
> > > > >       Avg: 21.229
> > > > >     90pct: 22.739
> > > > >       Max: 30.491
> > > > > root@cerowrt:~# /usr/lib/CeroWrtScripts/netperfrunner.sh
> > > > > 2014-10-19 15:11:54 Testing netperf.bufferbloat.net (ipv4) with 4 streams down a                                                                                                                                                             nd up while pinging gstatic.com. Takes about 60 seconds.
> > > > >  Download:  7.02 Mbps
> > > > >    Upload:  63.68 Mbps
> > > > >   Latency: (in msec, 61 pings, 0.00% packet loss)
> > > > >       Min: 21.690
> > > > >     10pct: 117.997
> > > > >    Median: 153.983
> > > > >       Avg: 150.851
> > > > >     90pct: 180.777
> > > > >       Max: 190.265
> > > > >
> > > > >
> > > > > The first result i gotten by only shaping the uplink and the second results are for only shaping the downlink...
> > > > >
> > > > > and for the testing i put 0 on the for the downlink and uplink when I was testing for the opposite. and I put 50000 in it when I was shaping it for the test. Am I doing it right ?
> > > > >
> > > > > On Sun, Oct 19, 2014 at 2:51 PM, Dave Taht <dave.taht@gmail.com> wrote:
> > > > > shape the download only, then measure. shape the upload only, then measure.
> > > > >
> > > > > On Sun, Oct 19, 2014 at 11:48 AM, Ernesto Elias <ernestogelias@gmail.com> wrote:
> > > > > > Oh OK I mean I was just wondering because I was trying to figure out what
> > > > > > should I have put in the sqm tab because I saw that the wndr3800 tops out at
> > > > > > 50 mbit combined of download and upload right? As of now I have it at 25/25
> > > > > > on the page. Is that right or should I just shape the upload?
> > > > > >
> > > > > > On Oct 19, 2014 2:24 PM, "Dave Taht" <dave.taht@gmail.com> wrote:
> > > > > >>
> > > > > >> On at least one verizon device I've tried it appeared that they had
> > > > > >> SFQ or something similar on egress from the modem.
> > > > > >>
> > > > > >>
> > > > > >> http://www.bufferbloat.net/projects/codel/wiki/RRUL_Rogues_Gallery#Verizon-FIOS-Testing-at-25Mbit-up-and-25Mbit-down
> > > > > >>
> > > > > >> So you only needed to shape the download. which is good as we start
> > > > > >> peaking out at 50Mbit download total. But only measurements can tell.
> > > > > >>
> > > > > >>
> > > > > >> On Sun, Oct 19, 2014 at 10:51 AM, Ernesto Elias <ernestogelias@gmail.com>
> > > > > >> wrote:
> > > > > >> > Hello everyone!
> > > > > >> > I have a question about the wndr3800 routing limit. I went back to the
> > > > > >> > older
> > > > > >> > submissions to see if I can find what would be the answer for it. But in
> > > > > >> > my
> > > > > >> > search I haven't managed to find a definite answer. From what I seen
> > > > > >> > about
> > > > > >> > setting the limit it can do with SQM is 50, 60, or 80 mbit. I'm just
> > > > > >> > wondering if anyone can shed some light for me here as I have verizon
> > > > > >> > fios
> > > > > >> > and my speeds are 50 dl/50 ul. Thank you guys very much!
> > > > > >> >
> > > > > >> > _______________________________________________
> > > > > >> > Cerowrt-devel mailing list
> > > > > >> > Cerowrt-devel@lists.bufferbloat.net
> > > > > >> > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > > > >> >
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >> --
> > > > > >> Dave Täht
> > > > > >>
> > > > > >> thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > > >
> > > > >
> > > > >
> > > > > --
> > > > > Dave Täht
> > > > >
> > > > > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks
> > > > >
> > > > > _______________________________________________
> > > > > Cerowrt-devel mailing list
> > > > > Cerowrt-devel@lists.bufferbloat.net
> > > > > https://lists.bufferbloat.net/listinfo/cerowrt-devel
> > > >
> > > >
> > >
> >
> 
> 


^ permalink raw reply	[flat|nested] 24+ messages in thread

end of thread, other threads:[~2014-10-24 11:38 UTC | newest]

Thread overview: 24+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-10-19 17:51 [Cerowrt-devel] Routing limit question Ernesto Elias
2014-10-19 18:24 ` Dave Taht
2014-10-19 18:48   ` Ernesto Elias
2014-10-19 18:51     ` Dave Taht
2014-10-19 19:16       ` Ernesto Elias
2014-10-19 19:55         ` Sebastian Moeller
2014-10-20  1:08           ` Ernesto Elias
2014-10-20  8:11             ` Sebastian Moeller
2014-10-21  2:30               ` Ernesto Elias
2014-10-21  8:23                 ` Sebastian Moeller
2014-10-22 19:39                   ` Ernesto Elias
2014-10-22 20:36                     ` Sebastian Moeller
2014-10-23 22:03                       ` Ernesto Elias
2014-10-24 11:38                         ` Sebastian Moeller
2014-10-19 18:56   ` Sebastian Moeller
2014-10-19 19:22     ` Dave Taht
2014-10-19 19:33       ` Ernesto Elias
2014-10-19 19:41         ` Dave Taht
     [not found]     ` <CAA93jw6DFxoDaRVCoQn=zxbOjgQ-dA1bCZ1W=p9br25jWrRgkg@mail.gmail.com>
2014-10-19 22:44       ` Sebastian Moeller
2014-10-20 20:04         ` Dave Taht
2014-10-20 22:48           ` Dave Taht
2014-10-21  3:35             ` Aaron Wood
2014-10-21  7:43               ` Sebastian Moeller
2014-10-21  7:51             ` Sebastian Moeller

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox