[Bloat] 22 seconds til bloat on gfiber?

[Bloat] 22 seconds til bloat on gfiber?

[Dave Taht]

http://www.dslreports.com/speedtest/5408767

-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jonathan Morton]

> On 23 Oct, 2016, at 00:56, Dave Taht <dave.taht@gmail.com> wrote:
> 
> http://www.dslreports.com/speedtest/5408767

Looks like that’s how long it takes for the throughput to ramp up to link capacity.  That in turn is a function of the sender’s TCP.

 - Jonathan Morton

Re: [Bloat] 22 seconds til bloat on gfiber?

[jb]

[-- Attachment #1: Type: text/plain, Size: 1231 bytes --]

This example takes about 6 seconds to get all the uploads running as
they are staged, and then each upload takes a while to get to full speed
because that is a function of  the senders TCP stack. So the smoothed
total transfer rate lags as well, and the whole thing doesn't start to bloat
out until we get to max speed.

There is an upload duration preference that can increase the total time
upload or download takes but people already have no patience and
close the tab when they start seeing decent upload numbers,
so increasing it just makes the quit rate higher still. For the quitters
we get no results at all, other than they quit before the end of the test.

thanks

On Sun, Oct 23, 2016 at 10:52 AM, Jonathan Morton <chromatix99@gmail.com>
wrote:

>
> > On 23 Oct, 2016, at 00:56, Dave Taht <dave.taht@gmail.com> wrote:
> >
> > http://www.dslreports.com/speedtest/5408767
>
> Looks like that’s how long it takes for the throughput to ramp up to link
> capacity.  That in turn is a function of the sender’s TCP.
>
>  - Jonathan Morton
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>

[-- Attachment #2: Type: text/html, Size: 2077 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Dave Taht]

On Sat, Oct 22, 2016 at 6:33 PM, jb <justin@dslr.net> wrote:
> This example takes about 6 seconds to get all the uploads running as
> they are staged, and then each upload takes a while to get to full speed
> because that is a function of  the senders TCP stack. So the smoothed
> total transfer rate lags as well, and the whole thing doesn't start to bloat
> out until we get to max speed.
>
> There is an upload duration preference that can increase the total time
> upload or download takes but people already have no patience and
> close the tab when they start seeing decent upload numbers,
> so increasing it just makes the quit rate higher still. For the quitters
> we get no results at all, other than they quit before the end of the test.

I agree that waiting that long is hard on users, and that since it
takes so long to get to that point, it will take a lot of work for a
gfiber user to stress out the connection, on a benchmark... but in the
real world, with a few users on the link, not so much.

400-1000ms latency when loaded counts as an "F" grade, in my opinion.
Perhaps doing the grade calculation only when the link is observed
near max bandwidth achieved (say, half)?

There are of course, other possible reasons for such bloat, like the
browser falling over, I wish I had a gfiber network and routing device
to test against.

Is there any way to browse
http://www.dslreports.com/speedtest/results/isp/r3910-google-fiber for
like the last 20 results to see if this is a common behavior on gfiber
for longer tests?

> thanks
>
> On Sun, Oct 23, 2016 at 10:52 AM, Jonathan Morton <chromatix99@gmail.com>
> wrote:
>>
>>
>> > On 23 Oct, 2016, at 00:56, Dave Taht <dave.taht@gmail.com> wrote:
>> >
>> > http://www.dslreports.com/speedtest/5408767
>>
>> Looks like that’s how long it takes for the throughput to ramp up to link
>> capacity.  That in turn is a function of the sender’s TCP.
>>
>>  - Jonathan Morton
>>
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>
>
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Bloat] 22 seconds til bloat on gfiber?

[Dave Taht]

randomly clicking around, 18 seconds to "start of bloat" on xfinity
http://www.dslreports.com/speedtest/5414347

On Sat, Oct 22, 2016 at 6:45 PM, Dave Taht <dave.taht@gmail.com> wrote:
> On Sat, Oct 22, 2016 at 6:33 PM, jb <justin@dslr.net> wrote:
>> This example takes about 6 seconds to get all the uploads running as
>> they are staged, and then each upload takes a while to get to full speed
>> because that is a function of  the senders TCP stack. So the smoothed
>> total transfer rate lags as well, and the whole thing doesn't start to bloat
>> out until we get to max speed.
>>
>> There is an upload duration preference that can increase the total time
>> upload or download takes but people already have no patience and
>> close the tab when they start seeing decent upload numbers,
>> so increasing it just makes the quit rate higher still. For the quitters
>> we get no results at all, other than they quit before the end of the test.
>
> I agree that waiting that long is hard on users, and that since it
> takes so long to get to that point, it will take a lot of work for a
> gfiber user to stress out the connection, on a benchmark... but in the
> real world, with a few users on the link, not so much.
>
> 400-1000ms latency when loaded counts as an "F" grade, in my opinion.
> Perhaps doing the grade calculation only when the link is observed
> near max bandwidth achieved (say, half)?
>
> There are of course, other possible reasons for such bloat, like the
> browser falling over, I wish I had a gfiber network and routing device
> to test against.
>
> Is there any way to browse
> http://www.dslreports.com/speedtest/results/isp/r3910-google-fiber for
> like the last 20 results to see if this is a common behavior on gfiber
> for longer tests?
>
>> thanks
>>
>> On Sun, Oct 23, 2016 at 10:52 AM, Jonathan Morton <chromatix99@gmail.com>
>> wrote:
>>>
>>>
>>> > On 23 Oct, 2016, at 00:56, Dave Taht <dave.taht@gmail.com> wrote:
>>> >
>>> > http://www.dslreports.com/speedtest/5408767
>>>
>>> Looks like that’s how long it takes for the throughput to ramp up to link
>>> capacity.  That in turn is a function of the sender’s TCP.
>>>
>>>  - Jonathan Morton
>>>
>>> _______________________________________________
>>> Bloat mailing list
>>> Bloat@lists.bufferbloat.net
>>> https://lists.bufferbloat.net/listinfo/bloat
>>
>>
>>
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>>
>
>
>
> --
> Dave Täht
> Let's go make home routers and wifi faster! With better software!
> http://blog.cerowrt.org



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Bloat] 22 seconds til bloat on gfiber?

[jb]

[-- Attachment #1: Type: text/plain, Size: 3320 bytes --]

Bloat starts here half way through upload, but it isn't nearly as bad as
the original example:
  https://www.dslreports.com/speedtest/5409732

I wonder if there is a browser or PC component to bad results on google
fiber I imagine they all have the same equipment?


On Sun, Oct 23, 2016 at 12:47 PM, Dave Taht <dave.taht@gmail.com> wrote:

> randomly clicking around, 18 seconds to "start of bloat" on xfinity
> http://www.dslreports.com/speedtest/5414347
>
> On Sat, Oct 22, 2016 at 6:45 PM, Dave Taht <dave.taht@gmail.com> wrote:
> > On Sat, Oct 22, 2016 at 6:33 PM, jb <justin@dslr.net> wrote:
> >> This example takes about 6 seconds to get all the uploads running as
> >> they are staged, and then each upload takes a while to get to full speed
> >> because that is a function of  the senders TCP stack. So the smoothed
> >> total transfer rate lags as well, and the whole thing doesn't start to
> bloat
> >> out until we get to max speed.
> >>
> >> There is an upload duration preference that can increase the total time
> >> upload or download takes but people already have no patience and
> >> close the tab when they start seeing decent upload numbers,
> >> so increasing it just makes the quit rate higher still. For the quitters
> >> we get no results at all, other than they quit before the end of the
> test.
> >
> > I agree that waiting that long is hard on users, and that since it
> > takes so long to get to that point, it will take a lot of work for a
> > gfiber user to stress out the connection, on a benchmark... but in the
> > real world, with a few users on the link, not so much.
> >
> > 400-1000ms latency when loaded counts as an "F" grade, in my opinion.
> > Perhaps doing the grade calculation only when the link is observed
> > near max bandwidth achieved (say, half)?
> >
> > There are of course, other possible reasons for such bloat, like the
> > browser falling over, I wish I had a gfiber network and routing device
> > to test against.
> >
> > Is there any way to browse
> > http://www.dslreports.com/speedtest/results/isp/r3910-google-fiber for
> > like the last 20 results to see if this is a common behavior on gfiber
> > for longer tests?
> >
> >> thanks
> >>
> >> On Sun, Oct 23, 2016 at 10:52 AM, Jonathan Morton <
> chromatix99@gmail.com>
> >> wrote:
> >>>
> >>>
> >>> > On 23 Oct, 2016, at 00:56, Dave Taht <dave.taht@gmail.com> wrote:
> >>> >
> >>> > http://www.dslreports.com/speedtest/5408767
> >>>
> >>> Looks like that’s how long it takes for the throughput to ramp up to
> link
> >>> capacity.  That in turn is a function of the sender’s TCP.
> >>>
> >>>  - Jonathan Morton
> >>>
> >>> _______________________________________________
> >>> Bloat mailing list
> >>> Bloat@lists.bufferbloat.net
> >>> https://lists.bufferbloat.net/listinfo/bloat
> >>
> >>
> >>
> >> _______________________________________________
> >> Bloat mailing list
> >> Bloat@lists.bufferbloat.net
> >> https://lists.bufferbloat.net/listinfo/bloat
> >>
> >
> >
> >
> > --
> > Dave Täht
> > Let's go make home routers and wifi faster! With better software!
> > http://blog.cerowrt.org
>
>
>
> --
> Dave Täht
> Let's go make home routers and wifi faster! With better software!
> http://blog.cerowrt.org
>

[-- Attachment #2: Type: text/html, Size: 5252 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Benjamin Cronce]

[-- Attachment #1: Type: text/plain, Size: 3716 bytes --]

On the opposite side of things. I found these. I wish more people did high
resolution samples.
http://www.dslreports.com/speedtest/5414499
http://www.dslreports.com/speedtest/5414506

Why does Google Fiber have so much bloat? They're running line rate. This
means their buffers are actually sized to have over 1 second of data at
line rate. I understand the underlying protocol is encapsulating groups of
Ethernet packets, which increases the burstiness in which the packets are
dequeued, but that's insane.

On Sat, Oct 22, 2016 at 8:47 PM, Dave Taht <dave.taht@gmail.com> wrote:

> randomly clicking around, 18 seconds to "start of bloat" on xfinity
> http://www.dslreports.com/speedtest/5414347
>
> On Sat, Oct 22, 2016 at 6:45 PM, Dave Taht <dave.taht@gmail.com> wrote:
> > On Sat, Oct 22, 2016 at 6:33 PM, jb <justin@dslr.net> wrote:
> >> This example takes about 6 seconds to get all the uploads running as
> >> they are staged, and then each upload takes a while to get to full speed
> >> because that is a function of  the senders TCP stack. So the smoothed
> >> total transfer rate lags as well, and the whole thing doesn't start to
> bloat
> >> out until we get to max speed.
> >>
> >> There is an upload duration preference that can increase the total time
> >> upload or download takes but people already have no patience and
> >> close the tab when they start seeing decent upload numbers,
> >> so increasing it just makes the quit rate higher still. For the quitters
> >> we get no results at all, other than they quit before the end of the
> test.
> >
> > I agree that waiting that long is hard on users, and that since it
> > takes so long to get to that point, it will take a lot of work for a
> > gfiber user to stress out the connection, on a benchmark... but in the
> > real world, with a few users on the link, not so much.
> >
> > 400-1000ms latency when loaded counts as an "F" grade, in my opinion.
> > Perhaps doing the grade calculation only when the link is observed
> > near max bandwidth achieved (say, half)?
> >
> > There are of course, other possible reasons for such bloat, like the
> > browser falling over, I wish I had a gfiber network and routing device
> > to test against.
> >
> > Is there any way to browse
> > http://www.dslreports.com/speedtest/results/isp/r3910-google-fiber for
> > like the last 20 results to see if this is a common behavior on gfiber
> > for longer tests?
> >
> >> thanks
> >>
> >> On Sun, Oct 23, 2016 at 10:52 AM, Jonathan Morton <
> chromatix99@gmail.com>
> >> wrote:
> >>>
> >>>
> >>> > On 23 Oct, 2016, at 00:56, Dave Taht <dave.taht@gmail.com> wrote:
> >>> >
> >>> > http://www.dslreports.com/speedtest/5408767
> >>>
> >>> Looks like that’s how long it takes for the throughput to ramp up to
> link
> >>> capacity.  That in turn is a function of the sender’s TCP.
> >>>
> >>>  - Jonathan Morton
> >>>
> >>> _______________________________________________
> >>> Bloat mailing list
> >>> Bloat@lists.bufferbloat.net
> >>> https://lists.bufferbloat.net/listinfo/bloat
> >>
> >>
> >>
> >> _______________________________________________
> >> Bloat mailing list
> >> Bloat@lists.bufferbloat.net
> >> https://lists.bufferbloat.net/listinfo/bloat
> >>
> >
> >
> >
> > --
> > Dave Täht
> > Let's go make home routers and wifi faster! With better software!
> > http://blog.cerowrt.org
>
>
>
> --
> Dave Täht
> Let's go make home routers and wifi faster! With better software!
> http://blog.cerowrt.org
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>

[-- Attachment #2: Type: text/html, Size: 5868 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Mikael Abrahamsson]

On Sat, 22 Oct 2016, Dave Taht wrote:

> http://www.dslreports.com/speedtest/5408767

What's the setup here? Someone has told me that Google Fiber is PON? So 
there is an ONT at the customer prem which takes fiber and hands off some 
kind of 1GBASE-T? What more?

Just trying to figure out what device has ~13-15 megabyte buffer so it can 
induce 1200ms buffer lag at 1 gigabit/s.

-- 
Mikael Abrahamsson    email: swmike@swm.pp.se

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jonathan Morton]

[-- Attachment #1: Type: text/plain, Size: 153 bytes --]

Surely it's attached via Gigabit Ethernet?  The bloat could easily be in
the NIC or its drivers.  That has nothing to do with Google.

- Jonathan Morton

[-- Attachment #2: Type: text/html, Size: 201 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[jb]

[-- Attachment #1: Type: text/plain, Size: 486 bytes --]

I've adjusted the display to list up to 10 of the last registered users and
will contact some of them to see if I can get someone to run some hi-res
tests, confirm with a command line ping, and offer some insight into their
hardware setup..


On Sun, Oct 23, 2016 at 7:29 PM, Jonathan Morton <chromatix99@gmail.com>
wrote:

> Surely it's attached via Gigabit Ethernet?  The bloat could easily be in
> the NIC or its drivers.  That has nothing to do with Google.
>
> - Jonathan Morton
>

[-- Attachment #2: Type: text/html, Size: 896 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Klatsky, Carl]

[-- Attachment #1: Type: text/plain, Size: 879 bytes --]

Justin,

How does one initiate the ‘hi-res’ tests?  On the site, I am only seeing the regular speed / bloat test get launched.

Side comment – could you please label the graph with units?  I read it as X-axis is time in seconds and Y-axis as latency in milliseconds, but having the explicit label will be helpful.  Thanks.

Regards,
Carl Klatsky

I've adjusted the display to list up to 10 of the last registered users and will contact some of them to see if I can get someone to run some hi-res tests, confirm with a command line ping, and offer some insight into their hardware setup..


On Sun, Oct 23, 2016 at 7:29 PM, Jonathan Morton <chromatix99@gmail.com<mailto:chromatix99@gmail.com>> wrote:

Surely it's attached via Gigabit Ethernet?  The bloat could easily be in the NIC or its drivers.  That has nothing to do with Google.

- Jonathan Morton


[-- Attachment #2: Type: text/html, Size: 4518 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[jb]

[-- Attachment #1.1: Type: text/plain, Size: 1166 bytes --]

Sure, will take a look at the graph.

In the test preferences screen, under Advanced.
see screen shot hopefully attached.

thanks.

On Mon, Oct 24, 2016 at 3:14 AM, Klatsky, Carl <Carl_Klatsky@comcast.com>
wrote:

> Justin,
>
>
>
> How does one initiate the ‘hi-res’ tests?  On the site, I am only seeing
> the regular speed / bloat test get launched.
>
>
>
> Side comment – could you please label the graph with units?  I read it as
> X-axis is time in seconds and Y-axis as latency in milliseconds, but having
> the explicit label will be helpful.  Thanks.
>
>
>
> Regards,
>
> Carl Klatsky
>
>
>
> I've adjusted the display to list up to 10 of the last registered users
> and will contact some of them to see if I can get someone to run some
> hi-res tests, confirm with a command line ping, and offer some insight into
> their hardware setup..
>
>
>
>
>
> On Sun, Oct 23, 2016 at 7:29 PM, Jonathan Morton <chromatix99@gmail.com>
> wrote:
>
> Surely it's attached via Gigabit Ethernet?  The bloat could easily be in
> the NIC or its drivers.  That has nothing to do with Google.
>
> - Jonathan Morton
>
>
>

[-- Attachment #1.2: Type: text/html, Size: 3431 bytes --]

[-- Attachment #2: Screen Shot 2016-10-24 at 11.44.50 am.png --]
[-- Type: image/png, Size: 64534 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Klatsky, Carl]

[-- Attachment #1: Type: text/plain, Size: 1432 bytes --]

Thanks Justin.  I found the Preferences and set to Hi-Res bufferbloat testing, but I don’t see the Hi-Res when I click the graph in my test results.  The Hi-Res test should run longer than the standard speed test, yes?

http://www.dslreports.com/speedtest/5428417

Regards,
Carl Klatsky

Sure, will take a look at the graph.

In the test preferences screen, under Advanced.
see screen shot hopefully attached.

thanks.

On Mon, Oct 24, 2016 at 3:14 AM, Klatsky, Carl <Carl_Klatsky@comcast.com<mailto:Carl_Klatsky@comcast.com>> wrote:
Justin,

How does one initiate the ‘hi-res’ tests?  On the site, I am only seeing the regular speed / bloat test get launched.

Side comment – could you please label the graph with units?  I read it as X-axis is time in seconds and Y-axis as latency in milliseconds, but having the explicit label will be helpful.  Thanks.

Regards,
Carl Klatsky

I've adjusted the display to list up to 10 of the last registered users and will contact some of them to see if I can get someone to run some hi-res tests, confirm with a command line ping, and offer some insight into their hardware setup..


On Sun, Oct 23, 2016 at 7:29 PM, Jonathan Morton <chromatix99@gmail.com<mailto:chromatix99@gmail.com>> wrote:

Surely it's attached via Gigabit Ethernet?  The bloat could easily be in the NIC or its drivers.  That has nothing to do with Google.

- Jonathan Morton



[-- Attachment #2: Type: text/html, Size: 7433 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Noah Causin]

[-- Attachment #1: Type: text/plain, Size: 2127 bytes --]

I have been having an issue where enabling the Hi-Res bufferbloat 
setting doesn't work.

What I do is press the "Create" option at the bottom of the preferences 
menu.  I input the settings that match my internet connection, and when 
I go back to the speedtest page, I have that button to use.

Unfortunately, it makes all the other buttons go away.


On 10/24/2016 9:13 AM, Klatsky, Carl wrote:
>
> Thanks Justin.  I found the Preferences and set to Hi-Res bufferbloat 
> testing, but I don’t see the Hi-Res when I click the graph in my test 
> results.  The Hi-Res test should run longer than the standard speed 
> test, yes?
>
> http://www.dslreports.com/speedtest/5428417
>
> Regards,
>
> Carl Klatsky
>
> Sure, will take a look at the graph.
>
> In the test preferences screen, under Advanced.
>
> see screen shot hopefully attached.
>
> thanks.
>
> On Mon, Oct 24, 2016 at 3:14 AM, Klatsky, Carl 
> <Carl_Klatsky@comcast.com <mailto:Carl_Klatsky@comcast.com>> wrote:
>
>     Justin,
>
>     How does one initiate the ‘hi-res’ tests?  On the site, I am only
>     seeing the regular speed / bloat test get launched.
>
>     Side comment – could you please label the graph with units?  I
>     read it as X-axis is time in seconds and Y-axis as latency in
>     milliseconds, but having the explicit label will be helpful.  Thanks.
>
>     Regards,
>
>     Carl Klatsky
>
>     I've adjusted the display to list up to 10 of the last registered
>     users and will contact some of them to see if I can get someone to
>     run some hi-res tests, confirm with a command line ping, and offer
>     some insight into their hardware setup..
>
>     On Sun, Oct 23, 2016 at 7:29 PM, Jonathan Morton
>     <chromatix99@gmail.com <mailto:chromatix99@gmail.com>> wrote:
>
>         Surely it's attached via Gigabit Ethernet?  The bloat could
>         easily be in the NIC or its drivers.  That has nothing to do
>         with Google.
>
>         - Jonathan Morton
>
>
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat


[-- Attachment #2: Type: text/html, Size: 10470 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jonas Mårtensson]

[-- Attachment #1: Type: text/plain, Size: 917 bytes --]

On Sun, Oct 23, 2016 at 7:50 AM, Mikael Abrahamsson <swmike@swm.pp.se>
wrote:

> On Sat, 22 Oct 2016, Dave Taht wrote:
>
> http://www.dslreports.com/speedtest/5408767
>>
>
> What's the setup here? Someone has told me that Google Fiber is PON? So
> there is an ONT at the customer prem which takes fiber and hands off some
> kind of 1GBASE-T? What more?
>
> Just trying to figure out what device has ~13-15 megabyte buffer so it can
> induce 1200ms buffer lag at 1 gigabit/s.


Of course I don't know the setup in this particular case but from what I've
heard, gfiber is mostly GPON with 1:16 splits. They have used different ONT
versions but the "current" generation seems to be a custom ONT from an ODM:

https://support.google.com/fiber/answer/6035992?hl=en&ref_topic=2667450
https://wikidevi.com/wiki/Google_Fiber_Jack_v2_(GFLT110)

Also, this may be interesting:

https://gfiber-review.googlesource.com/#/c/1232/

[-- Attachment #2: Type: text/html, Size: 1861 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Klatsky, Carl]

[-- Attachment #1: Type: text/plain, Size: 2143 bytes --]

Thanks.  Tried those steps, but they did not work to generate the ‘hi-res’ test.

Regards,
Carl Klatsky

I have been having an issue where enabling the Hi-Res bufferbloat setting doesn't work.

What I do is press the "Create" option at the bottom of the preferences menu.  I input the settings that match my internet connection, and when I go back to the speedtest page, I have that button to use.

Unfortunately, it makes all the other buttons go away.

On 10/24/2016 9:13 AM, Klatsky, Carl wrote:
Thanks Justin.  I found the Preferences and set to Hi-Res bufferbloat testing, but I don’t see the Hi-Res when I click the graph in my test results.  The Hi-Res test should run longer than the standard speed test, yes?

http://www.dslreports.com/speedtest/5428417

Regards,
Carl Klatsky

Sure, will take a look at the graph.

In the test preferences screen, under Advanced.
see screen shot hopefully attached.

thanks.

On Mon, Oct 24, 2016 at 3:14 AM, Klatsky, Carl <Carl_Klatsky@comcast.com<mailto:Carl_Klatsky@comcast.com>> wrote:
Justin,

How does one initiate the ‘hi-res’ tests?  On the site, I am only seeing the regular speed / bloat test get launched.

Side comment – could you please label the graph with units?  I read it as X-axis is time in seconds and Y-axis as latency in milliseconds, but having the explicit label will be helpful.  Thanks.

Regards,
Carl Klatsky

I've adjusted the display to list up to 10 of the last registered users and will contact some of them to see if I can get someone to run some hi-res tests, confirm with a command line ping, and offer some insight into their hardware setup..


On Sun, Oct 23, 2016 at 7:29 PM, Jonathan Morton <chromatix99@gmail.com<mailto:chromatix99@gmail.com>> wrote:

Surely it's attached via Gigabit Ethernet?  The bloat could easily be in the NIC or its drivers.  That has nothing to do with Google.

- Jonathan Morton






_______________________________________________

Bloat mailing list

Bloat@lists.bufferbloat.net<mailto:Bloat@lists.bufferbloat.net>

https://lists.bufferbloat.net/listinfo/bloat


[-- Attachment #2: Type: text/html, Size: 9743 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Benjamin Cronce]

[-- Attachment #1: Type: text/plain, Size: 1334 bytes --]

WDM-PON, giving each customer their own lambda of bandwidth. Effectively a
1:1 split.

On Mon, Oct 24, 2016 at 11:57 AM, Jonas Mårtensson <
martensson.jonas@gmail.com> wrote:

> On Sun, Oct 23, 2016 at 7:50 AM, Mikael Abrahamsson <swmike@swm.pp.se>
> wrote:
>
>> On Sat, 22 Oct 2016, Dave Taht wrote:
>>
>> http://www.dslreports.com/speedtest/5408767
>>>
>>
>> What's the setup here? Someone has told me that Google Fiber is PON? So
>> there is an ONT at the customer prem which takes fiber and hands off some
>> kind of 1GBASE-T? What more?
>>
>> Just trying to figure out what device has ~13-15 megabyte buffer so it
>> can induce 1200ms buffer lag at 1 gigabit/s.
>
>
> Of course I don't know the setup in this particular case but from what
> I've heard, gfiber is mostly GPON with 1:16 splits. They have used
> different ONT versions but the "current" generation seems to be a custom
> ONT from an ODM:
>
> https://support.google.com/fiber/answer/6035992?hl=en&ref_topic=2667450
> https://wikidevi.com/wiki/Google_Fiber_Jack_v2_(GFLT110)
>
> Also, this may be interesting:
>
> https://gfiber-review.googlesource.com/#/c/1232/
>
>
>
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>
>

[-- Attachment #2: Type: text/html, Size: 2747 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[jb]

[-- Attachment #1: Type: text/plain, Size: 2458 bytes --]

You should not need to create new buttons, I set bloat high frequency as an
anonymous user in preferences, saved the prefs, ran the test - but as http
- and it worked. However bloat high frequency was auto-disabled for https
because I was unsure it was valid to do high frequency pinging over SSL.

Here is a result of turning it on:
    https://www.dslreports.com/speedtest/5435977
you can see in the bloat graph, the difference in ping resolution.

However the test duration is not changed, if you want to run for longer
you should increase the duration parameters.

I've disabled the disabling of bloat high frequency over https so give
it another go. first, go to preferences and verify the checkbox is checked.
then run it. Send me a link to the results if you don't see any difference
still.

thanks
-Justin


On Tue, Oct 25, 2016 at 9:10 AM, Benjamin Cronce <bcronce@gmail.com> wrote:

> WDM-PON, giving each customer their own lambda of bandwidth. Effectively a
> 1:1 split.
>
> On Mon, Oct 24, 2016 at 11:57 AM, Jonas Mårtensson <
> martensson.jonas@gmail.com> wrote:
>
>> On Sun, Oct 23, 2016 at 7:50 AM, Mikael Abrahamsson <swmike@swm.pp.se>
>> wrote:
>>
>>> On Sat, 22 Oct 2016, Dave Taht wrote:
>>>
>>> http://www.dslreports.com/speedtest/5408767
>>>>
>>>
>>> What's the setup here? Someone has told me that Google Fiber is PON? So
>>> there is an ONT at the customer prem which takes fiber and hands off some
>>> kind of 1GBASE-T? What more?
>>>
>>> Just trying to figure out what device has ~13-15 megabyte buffer so it
>>> can induce 1200ms buffer lag at 1 gigabit/s.
>>
>>
>> Of course I don't know the setup in this particular case but from what
>> I've heard, gfiber is mostly GPON with 1:16 splits. They have used
>> different ONT versions but the "current" generation seems to be a custom
>> ONT from an ODM:
>>
>> https://support.google.com/fiber/answer/6035992?hl=en&ref_topic=2667450
>> https://wikidevi.com/wiki/Google_Fiber_Jack_v2_(GFLT110)
>>
>> Also, this may be interesting:
>>
>> https://gfiber-review.googlesource.com/#/c/1232/
>>
>>
>>
>>
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>>
>>
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>
>

[-- Attachment #2: Type: text/html, Size: 4601 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jonas Mårtensson]

[-- Attachment #1: Type: text/plain, Size: 463 bytes --]

On Sun, Oct 23, 2016 at 1:52 AM, Jonathan Morton <chromatix99@gmail.com>
wrote:

>
> > On 23 Oct, 2016, at 00:56, Dave Taht <dave.taht@gmail.com> wrote:
> >
> > http://www.dslreports.com/speedtest/5408767
>
> Looks like that’s how long it takes for the throughput to ramp up to link
> capacity.  That in turn is a function of the sender’s TCP.


In this case the slow ramp up may also be related to the DBA algorithm used
in the GPON equipment.

[-- Attachment #2: Type: text/html, Size: 933 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Benjamin Cronce]

[-- Attachment #1: Type: text/plain, Size: 2688 bytes --]

Is there a max duration*bandwidth we should use to be courteous?

On Mon, Oct 24, 2016 at 7:09 PM, jb <justin@dslr.net> wrote:

> You should not need to create new buttons, I set bloat high frequency as
> an anonymous user in preferences, saved the prefs, ran the test - but as
> http - and it worked. However bloat high frequency was auto-disabled for
> https because I was unsure it was valid to do high frequency pinging over
> SSL.
>
> Here is a result of turning it on:
>     https://www.dslreports.com/speedtest/5435977
> you can see in the bloat graph, the difference in ping resolution.
>
> However the test duration is not changed, if you want to run for longer
> you should increase the duration parameters.
>
> I've disabled the disabling of bloat high frequency over https so give
> it another go. first, go to preferences and verify the checkbox is checked.
> then run it. Send me a link to the results if you don't see any difference
> still.
>
> thanks
> -Justin
>
>
> On Tue, Oct 25, 2016 at 9:10 AM, Benjamin Cronce <bcronce@gmail.com>
> wrote:
>
>> WDM-PON, giving each customer their own lambda of bandwidth. Effectively
>> a 1:1 split.
>>
>> On Mon, Oct 24, 2016 at 11:57 AM, Jonas Mårtensson <
>> martensson.jonas@gmail.com> wrote:
>>
>>> On Sun, Oct 23, 2016 at 7:50 AM, Mikael Abrahamsson <swmike@swm.pp.se>
>>> wrote:
>>>
>>>> On Sat, 22 Oct 2016, Dave Taht wrote:
>>>>
>>>> http://www.dslreports.com/speedtest/5408767
>>>>>
>>>>
>>>> What's the setup here? Someone has told me that Google Fiber is PON? So
>>>> there is an ONT at the customer prem which takes fiber and hands off some
>>>> kind of 1GBASE-T? What more?
>>>>
>>>> Just trying to figure out what device has ~13-15 megabyte buffer so it
>>>> can induce 1200ms buffer lag at 1 gigabit/s.
>>>
>>>
>>> Of course I don't know the setup in this particular case but from what
>>> I've heard, gfiber is mostly GPON with 1:16 splits. They have used
>>> different ONT versions but the "current" generation seems to be a custom
>>> ONT from an ODM:
>>>
>>> https://support.google.com/fiber/answer/6035992?hl=en&ref_topic=2667450
>>> https://wikidevi.com/wiki/Google_Fiber_Jack_v2_(GFLT110)
>>>
>>> Also, this may be interesting:
>>>
>>> https://gfiber-review.googlesource.com/#/c/1232/
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Bloat mailing list
>>> Bloat@lists.bufferbloat.net
>>> https://lists.bufferbloat.net/listinfo/bloat
>>>
>>>
>>
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>>
>>
>

[-- Attachment #2: Type: text/html, Size: 5135 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Klatsky, Carl]

[-- Attachment #1: Type: text/plain, Size: 3153 bytes --]

Perhaps I am misunderstanding the different output between the ‘low-res’ and ‘high-res’ versions of the test.  The link below was from Corporate LAN with the ‘hi-res’ preference set & saved.  To me the output looks like prior test run output.  Maybe I have been seeing ‘hi-res’ output all along and don’t have a comparison to ‘low-res’ output?

http://www.dslreports.com/speedtest/5440827

Regards,
Carl Klatsky

From: Bloat [mailto:bloat-bounces@lists.bufferbloat.net] On Behalf Of jb
Sent: Monday, October 24, 2016 8:10 PM
To: Benjamin Cronce <bcronce@gmail.com>
Cc: bloat <bloat@lists.bufferbloat.net>
Subject: Re: [Bloat] 22 seconds til bloat on gfiber?

You should not need to create new buttons, I set bloat high frequency as an anonymous user in preferences, saved the prefs, ran the test - but as http - and it worked. However bloat high frequency was auto-disabled for https because I was unsure it was valid to do high frequency pinging over SSL.

Here is a result of turning it on:
    https://www.dslreports.com/speedtest/5435977
you can see in the bloat graph, the difference in ping resolution.

However the test duration is not changed, if you want to run for longer
you should increase the duration parameters.

I've disabled the disabling of bloat high frequency over https so give
it another go. first, go to preferences and verify the checkbox is checked.
then run it. Send me a link to the results if you don't see any difference
still.

thanks
-Justin


On Tue, Oct 25, 2016 at 9:10 AM, Benjamin Cronce <bcronce@gmail.com<mailto:bcronce@gmail.com>> wrote:
WDM-PON, giving each customer their own lambda of bandwidth. Effectively a 1:1 split.

On Mon, Oct 24, 2016 at 11:57 AM, Jonas Mårtensson <martensson.jonas@gmail.com<mailto:martensson.jonas@gmail.com>> wrote:
On Sun, Oct 23, 2016 at 7:50 AM, Mikael Abrahamsson <swmike@swm.pp.se<mailto:swmike@swm.pp.se>> wrote:
On Sat, 22 Oct 2016, Dave Taht wrote:
http://www.dslreports.com/speedtest/5408767

What's the setup here? Someone has told me that Google Fiber is PON? So there is an ONT at the customer prem which takes fiber and hands off some kind of 1GBASE-T? What more?

Just trying to figure out what device has ~13-15 megabyte buffer so it can induce 1200ms buffer lag at 1 gigabit/s.

Of course I don't know the setup in this particular case but from what I've heard, gfiber is mostly GPON with 1:16 splits. They have used different ONT versions but the "current" generation seems to be a custom ONT from an ODM:

https://support.google.com/fiber/answer/6035992?hl=en&ref_topic=2667450
https://wikidevi.com/wiki/Google_Fiber_Jack_v2_(GFLT110)

Also, this may be interesting:

https://gfiber-review.googlesource.com/#/c/1232/




_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net<mailto:Bloat@lists.bufferbloat.net>
https://lists.bufferbloat.net/listinfo/bloat


_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net<mailto:Bloat@lists.bufferbloat.net>
https://lists.bufferbloat.net/listinfo/bloat


[-- Attachment #2: Type: text/html, Size: 9758 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jan Ceuleers]

On 25/10/16 00:10, Benjamin Cronce wrote:
> WDM-PON, giving each customer their own lambda of bandwidth. Effectively
> a 1:1 split.

Not quite. All it means is that multiple PONs coexist on the same
outside plant, each on a different wavelength, and each serving multiple
end-users. Allows for higher densities.

What you suggest could be done but would quickly become impractical.

Re: [Bloat] 22 seconds til bloat on gfiber?

[jb]

[-- Attachment #1: Type: text/plain, Size: 3959 bytes --]

Carl, that run you linked to didn't have bloat high frequency option set.

To set it, go to preferences, check the box in the advanced section, save,
go BACK to preferences to make sure it is set. Then run a test.

If it is set the bloat graph will not be one probe per second anymore.
I did post a screenshot but the bloat mailing list doesn't like "large" 40kb
attachments lol.

To answer the other question on duration - 30 seconds is the system imposed
limit, you can set it to 30 if you have the patience.

thanks
-Justin

On Wed, Oct 26, 2016 at 12:11 AM, Klatsky, Carl <Carl_Klatsky@comcast.com>
wrote:

> Perhaps I am misunderstanding the different output between the ‘low-res’
> and ‘high-res’ versions of the test.  The link below was from Corporate LAN
> with the ‘hi-res’ preference set & saved.  To me the output looks like
> prior test run output.  Maybe I have been seeing ‘hi-res’ output all along
> and don’t have a comparison to ‘low-res’ output?
>
>
>
> http://www.dslreports.com/speedtest/5440827
>
>
>
> Regards,
>
> Carl Klatsky
>
>
>
> *From:* Bloat [mailto:bloat-bounces@lists.bufferbloat.net] *On Behalf Of *
> jb
> *Sent:* Monday, October 24, 2016 8:10 PM
> *To:* Benjamin Cronce <bcronce@gmail.com>
> *Cc:* bloat <bloat@lists.bufferbloat.net>
> *Subject:* Re: [Bloat] 22 seconds til bloat on gfiber?
>
>
>
> You should not need to create new buttons, I set bloat high frequency as
> an anonymous user in preferences, saved the prefs, ran the test - but as
> http - and it worked. However bloat high frequency was auto-disabled for
> https because I was unsure it was valid to do high frequency pinging over
> SSL.
>
>
>
> Here is a result of turning it on:
>
>     https://www.dslreports.com/speedtest/5435977
>
> you can see in the bloat graph, the difference in ping resolution.
>
>
>
> However the test duration is not changed, if you want to run for longer
>
> you should increase the duration parameters.
>
>
>
> I've disabled the disabling of bloat high frequency over https so give
>
> it another go. first, go to preferences and verify the checkbox is checked.
>
> then run it. Send me a link to the results if you don't see any difference
>
> still.
>
>
>
> thanks
>
> -Justin
>
>
>
>
>
> On Tue, Oct 25, 2016 at 9:10 AM, Benjamin Cronce <bcronce@gmail.com>
> wrote:
>
> WDM-PON, giving each customer their own lambda of bandwidth. Effectively a
> 1:1 split.
>
>
>
> On Mon, Oct 24, 2016 at 11:57 AM, Jonas Mårtensson <
> martensson.jonas@gmail.com> wrote:
>
> On Sun, Oct 23, 2016 at 7:50 AM, Mikael Abrahamsson <swmike@swm.pp.se>
> wrote:
>
> On Sat, 22 Oct 2016, Dave Taht wrote:
>
> http://www.dslreports.com/speedtest/5408767
>
>
> What's the setup here? Someone has told me that Google Fiber is PON? So
> there is an ONT at the customer prem which takes fiber and hands off some
> kind of 1GBASE-T? What more?
>
> Just trying to figure out what device has ~13-15 megabyte buffer so it can
> induce 1200ms buffer lag at 1 gigabit/s.
>
>
>
> Of course I don't know the setup in this particular case but from what
> I've heard, gfiber is mostly GPON with 1:16 splits. They have used
> different ONT versions but the "current" generation seems to be a custom
> ONT from an ODM:
>
>
>
> https://support.google.com/fiber/answer/6035992?hl=en&ref_topic=2667450
>
> https://wikidevi.com/wiki/Google_Fiber_Jack_v2_(GFLT110)
>
>
>
> Also, this may be interesting:
>
>
>
> https://gfiber-review.googlesource.com/#/c/1232/
>
>
>
>
>
>
>
>
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>
>
>
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>
>
>

[-- Attachment #2: Type: text/html, Size: 9741 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Klatsky, Carl]

[-- Attachment #1: Type: text/plain, Size: 4521 bytes --]

Thanks Justin.  I had been setting the option in the preferences, so not sure why it was not working before.  I cleared my browser cache and did the steps again, and now I am seeing the ‘hi-res’ results for bufferbloat.

http://www.dslreports.com/speedtest/5448121

Regards,
Carl Klatsky

From: justinbeech@gmail.com [mailto:justinbeech@gmail.com] On Behalf Of jb
Sent: Tuesday, October 25, 2016 5:20 PM
To: Klatsky, Carl <Carl_Klatsky@cable.comcast.com>
Cc: Benjamin Cronce <bcronce@gmail.com>; bloat <bloat@lists.bufferbloat.net>
Subject: Re: [Bloat] 22 seconds til bloat on gfiber?

Carl, that run you linked to didn't have bloat high frequency option set.

To set it, go to preferences, check the box in the advanced section, save, go BACK to preferences to make sure it is set. Then run a test.

If it is set the bloat graph will not be one probe per second anymore.
I did post a screenshot but the bloat mailing list doesn't like "large" 40kb
attachments lol.

To answer the other question on duration - 30 seconds is the system imposed limit, you can set it to 30 if you have the patience.

thanks
-Justin

On Wed, Oct 26, 2016 at 12:11 AM, Klatsky, Carl <Carl_Klatsky@comcast.com<mailto:Carl_Klatsky@comcast.com>> wrote:
Perhaps I am misunderstanding the different output between the ‘low-res’ and ‘high-res’ versions of the test.  The link below was from Corporate LAN with the ‘hi-res’ preference set & saved.  To me the output looks like prior test run output.  Maybe I have been seeing ‘hi-res’ output all along and don’t have a comparison to ‘low-res’ output?

http://www.dslreports.com/speedtest/5440827

Regards,
Carl Klatsky

From: Bloat [mailto:bloat-bounces@lists.bufferbloat.net<mailto:bloat-bounces@lists.bufferbloat.net>] On Behalf Of jb
Sent: Monday, October 24, 2016 8:10 PM
To: Benjamin Cronce <bcronce@gmail.com<mailto:bcronce@gmail.com>>
Cc: bloat <bloat@lists.bufferbloat.net<mailto:bloat@lists.bufferbloat.net>>
Subject: Re: [Bloat] 22 seconds til bloat on gfiber?

You should not need to create new buttons, I set bloat high frequency as an anonymous user in preferences, saved the prefs, ran the test - but as http - and it worked. However bloat high frequency was auto-disabled for https because I was unsure it was valid to do high frequency pinging over SSL.

Here is a result of turning it on:
    https://www.dslreports.com/speedtest/5435977
you can see in the bloat graph, the difference in ping resolution.

However the test duration is not changed, if you want to run for longer
you should increase the duration parameters.

I've disabled the disabling of bloat high frequency over https so give
it another go. first, go to preferences and verify the checkbox is checked.
then run it. Send me a link to the results if you don't see any difference
still.

thanks
-Justin


On Tue, Oct 25, 2016 at 9:10 AM, Benjamin Cronce <bcronce@gmail.com<mailto:bcronce@gmail.com>> wrote:
WDM-PON, giving each customer their own lambda of bandwidth. Effectively a 1:1 split.

On Mon, Oct 24, 2016 at 11:57 AM, Jonas Mårtensson <martensson.jonas@gmail.com<mailto:martensson.jonas@gmail.com>> wrote:
On Sun, Oct 23, 2016 at 7:50 AM, Mikael Abrahamsson <swmike@swm.pp.se<mailto:swmike@swm.pp.se>> wrote:
On Sat, 22 Oct 2016, Dave Taht wrote:
http://www.dslreports.com/speedtest/5408767

What's the setup here? Someone has told me that Google Fiber is PON? So there is an ONT at the customer prem which takes fiber and hands off some kind of 1GBASE-T? What more?

Just trying to figure out what device has ~13-15 megabyte buffer so it can induce 1200ms buffer lag at 1 gigabit/s.

Of course I don't know the setup in this particular case but from what I've heard, gfiber is mostly GPON with 1:16 splits. They have used different ONT versions but the "current" generation seems to be a custom ONT from an ODM:

https://support.google.com/fiber/answer/6035992?hl=en&ref_topic=2667450
https://wikidevi.com/wiki/Google_Fiber_Jack_v2_(GFLT110)

Also, this may be interesting:

https://gfiber-review.googlesource.com/#/c/1232/




_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net<mailto:Bloat@lists.bufferbloat.net>
https://lists.bufferbloat.net/listinfo/bloat


_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net<mailto:Bloat@lists.bufferbloat.net>
https://lists.bufferbloat.net/listinfo/bloat



[-- Attachment #2: Type: text/html, Size: 16672 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Benjamin Cronce]

[-- Attachment #1: Type: text/plain, Size: 2537 bytes --]

Sorry to side track

1:1 split bandwidth wise, still a 1:16 or whatever fiber split. Each port
can handle 40Gb/s, which is 32 lambdas of 1.25Gb/s, each customer getting
their own lambda. The ONT can either be WDM-PON or GPON with an inline
filter. A Google Fiber engineer actually had this in his blog a long while
back, talking about their design and the "dedicated" aspect of an unshared
GPON. PON can only handle about a 32 split before the signal strength gets
too low toe be practical. If each group of customers shared a lambda, they
would need too many split or repeaters, which is more impractical.

According to Sonic.Net, infrastructure and transit only constitutes about
1%-2% of the cost of being an ISP. May as well pony up for the best
infrastructure to reduce operational costs, which is where the bulk of the
cost of being an ISP rests. Repeaters and excess splits increase
operational costs.

I'm not entirely sure which part you mean "impractical". I actually have a
dedicated self-healing fiber loop from my home to my ISP's CO. $52.74/mo
after taxes.Well... $21.09/mo for the promo, and unbundled. Once at the CO,
it plugs into a patch panel where it then feeds into a splitter and into a
GPON port, but is otherwise dedicated back to the CO. The main benefit of
PON is the incredibly high density chassis. Not to mention a single WDM-PON
port consumes about 2x the power of a single 1Gb/1Gb active Ethernet port,
but can handle up to 32 ONTs and 40Gb/40Gb of bandwidth. Very power
efficient.

Google Fiber goes the route of fiber huts, while they're not absolutely
required, it's probably a good idea when covering a large area. There is
some very high density fiber bundles you can purchase, 144 strands of fiber
in the thickness of a pencil. If you ran that out to the normal size of a
conduit, you're getting into the 100k strands per conduit range, assuming
perfect packing density.

On Tue, Oct 25, 2016 at 10:59 AM, Jan Ceuleers <jan.ceuleers@gmail.com>
wrote:

> On 25/10/16 00:10, Benjamin Cronce wrote:
> > WDM-PON, giving each customer their own lambda of bandwidth. Effectively
> > a 1:1 split.
>
> Not quite. All it means is that multiple PONs coexist on the same
> outside plant, each on a different wavelength, and each serving multiple
> end-users. Allows for higher densities.
>
> What you suggest could be done but would quickly become impractical.
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>

[-- Attachment #2: Type: text/html, Size: 3240 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jan Ceuleers]

On 26/10/16 03:05, Benjamin Cronce wrote:
> Sorry to side track
> 
> 1:1 split bandwidth wise, still a 1:16 or whatever fiber split. Each
> port can handle 40Gb/s, which is 32 lambdas of 1.25Gb/s, each customer
> getting their own lambda. The ONT can either be WDM-PON or GPON with an
> inline filter. A Google Fiber engineer actually had this in his blog a
> long while back, talking about their design and the "dedicated" aspect
> of an unshared GPON. PON can only handle about a 32 split before the
> signal strength gets too low toe be practical. If each group of
> customers shared a lambda, they would need too many split or repeaters,
> which is more impractical.

I am aware of GPON deployments with splitting factors of 64 and higher,
and these do not (at all) pose a problem with the optical power budget.

> I'm not entirely sure which part you mean "impractical".

What I mean is that the OLT optics become very expensive if you need to
support as many lambdas as you have customers. You'd furthermore need an
OLT port for much fewer customers (e.g. 1 port per 64 or 128 customers)
than the thousands you can support on a (shared) GPON port on a single
lambda.

Also on the ONT side there is the danger of spiraling cost in that I
don't think it is economically feasible yet to incorporate tunable
lasers in the CPE, certainly not across 64 or 128 lambdas, meaning that
you'd need as many hardware variants of the ONT as you have lambdas.

But maybe I'm wrong.

Re: [Bloat] 22 seconds til bloat on gfiber?

[Mikael Abrahamsson]

On Wed, 26 Oct 2016, Jan Ceuleers wrote:

> What I mean is that the OLT optics become very expensive if you need to
> support as many lambdas as you have customers. You'd furthermore need an
> OLT port for much fewer customers (e.g. 1 port per 64 or 128 customers)
> than the thousands you can support on a (shared) GPON port on a single
> lambda.

That only works if your customers don't use their Internet access very 
much. If they do, you're in trouble and have to rebuild.

In my market, we're now in the access speeds where 100/10 is on the lower 
end of access, and it's not uncommon for people to have 250, 500 or 1000 
downstream. If they then actually start using their bw then you'd have to 
rebuild to either go higher speed for some CPE (complicated and 
expensive), or rebuild to have smaller splitter domains.

I guess the answer depends a lot on your cost of labour.

-- 
Mikael Abrahamsson    email: swmike@swm.pp.se

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jonas Mårtensson]

[-- Attachment #1: Type: text/plain, Size: 1378 bytes --]

On Wed, Oct 26, 2016 at 5:36 PM, Mikael Abrahamsson <swmike@swm.pp.se>
wrote:

> On Wed, 26 Oct 2016, Jan Ceuleers wrote:
>
> What I mean is that the OLT optics become very expensive if you need to
>> support as many lambdas as you have customers. You'd furthermore need an
>> OLT port for much fewer customers (e.g. 1 port per 64 or 128 customers)
>> than the thousands you can support on a (shared) GPON port on a single
>> lambda.
>>
>
> That only works if your customers don't use their Internet access very
> much. If they do, you're in trouble and have to rebuild.
>

Yes, and the question then becomes: How much is "very much"? This can of
course be analyzed mathematically, which e.g. Google have done here:

http://research.google.com/pubs/pub44935.html


>
> In my market, we're now in the access speeds where 100/10 is on the lower
> end of access, and it's not uncommon for people to have 250, 500 or 1000
> downstream. If they then actually start using their bw then you'd have to
> rebuild to either go higher speed for some CPE (complicated and expensive),
> or rebuild to have smaller splitter domains.
>

The standard answer from PON proponents (I'm not one) is to upgrade
equipment, from GPON to XG-PON or NG-PON2. But upgrading hardware as
bandwidth demand increases is necessary whatever the technology - what's
important is the scalability of the solution.

[-- Attachment #2: Type: text/html, Size: 2188 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Dave Taht]

On Thu, Oct 27, 2016 at 7:32 AM, Jonas Mårtensson
<martensson.jonas@gmail.com> wrote:
>
>
> On Wed, Oct 26, 2016 at 5:36 PM, Mikael Abrahamsson <swmike@swm.pp.se>
> wrote:
>>
>> On Wed, 26 Oct 2016, Jan Ceuleers wrote:
>>
>>> What I mean is that the OLT optics become very expensive if you need to
>>> support as many lambdas as you have customers. You'd furthermore need an
>>> OLT port for much fewer customers (e.g. 1 port per 64 or 128 customers)
>>> than the thousands you can support on a (shared) GPON port on a single
>>> lambda.
>>
>>
>> That only works if your customers don't use their Internet access very
>> much. If they do, you're in trouble and have to rebuild.
>
>
> Yes, and the question then becomes: How much is "very much"? This can of
> course be analyzed mathematically, which e.g. Google have done here:

When you are on one side of an S curve, it's hard to see where it flattens out.

I've been meaning to research and write a piece called "have we
reached 'Peak Bandwidth'"? for a while now.

My thesis is that what users actually want is short RTTs for
interactive, once basic bandwidth needs are slaked, which starts to
happen once you crack the largest typical load (which these days is 4k
video streaming).

gbit fiber is *way* on the unneeded side of the demand curve for home users.


> http://research.google.com/pubs/pub44935.html

which kind of points out that you need business users to use it all up.

>>
>>
>> In my market, we're now in the access speeds where 100/10 is on the lower
>> end of access, and it's not uncommon for people to have 250, 500 or 1000
>> downstream. If they then actually start using their bw then you'd have to
>> rebuild to either go higher speed for some CPE (complicated and expensive),
>> or rebuild to have smaller splitter domains.

I am curious if studies exists of the actual consumption in typical
100mbit and above plans, vs 20Mbit, here and worldwide. Again, my
thesis is, aside from business (and bittorrent) users, your typical
250mbit plan would have very close to the same consumption as the
100mbit plan. They'd use up 250mbits for a couple hours a month, but
that's it.

I am increasingly convinced that without a killer application that requires it,
we've hit "peak bandwidth".

>
> The standard answer from PON proponents (I'm not one) is to upgrade
> equipment, from GPON to XG-PON or NG-PON2. But upgrading hardware as
> bandwidth demand increases is necessary whatever the technology - what's
> important is the scalability of the solution.
>
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

[Bloat] Sidebar on s-curves

[David Collier-Brown]

On 27/10/16 12:56 PM, Dave Taht wrote:
> When you are on one side of an S curve, it's hard to see where it flattens out.
>
> I've been meaning to research and write a piece called "have we
> reached 'Peak Bandwidth'"? for a while now.
>
> My thesis is that what users actually want is short RTTs for
> interactive, once basic bandwidth needs are slaked, which starts to
> happen once you crack the largest typical load (which these days is 4k
> video streaming).
>
> gbit fiber is *way* on the unneeded side of the demand curve for home users.
>
It's not an s-curve, but Neil Gunther had done some very interesting 
work on load/throughput curves, specifically including predicting the 
drop-off after reaching 100%. Look for the Universal Scalability Law.

http://www.perfdynamics.com/Manifesto/USLscalability.html

I suspect it might be a variant of the same problem.

--dave

-- 
David Collier-Brown,         | Always do right. This will gratify
System Programmer and Author | some people and astonish the rest
davecb@spamcop.net           |                      -- Mark Twain

Re: [Bloat] 22 seconds til bloat on gfiber?

[David Lang]

On Thu, 27 Oct 2016, Dave Taht wrote:

> I am curious if studies exists of the actual consumption in typical
> 100mbit and above plans, vs 20Mbit, here and worldwide. Again, my
> thesis is, aside from business (and bittorrent) users, your typical
> 250mbit plan would have very close to the same consumption as the
> 100mbit plan. They'd use up 250mbits for a couple hours a month, but
> that's it.
>
> I am increasingly convinced that without a killer application that requires it,
> we've hit "peak bandwidth".

You sound like my College Professor from the early 90's who said that the 
networks were now going to be so fast that there was no way that users would 
need all the available bandwidth, and that it was up to the students in the 
class to invent new uses :-)

Then the web happened.

any declaration of 'peak bandwidth' is only a temporary state.

David Lang

Re: [Bloat] 22 seconds til bloat on gfiber?

[Dave Taht]

On Thu, Oct 27, 2016 at 12:30 PM, David Lang <david@lang.hm> wrote:
> On Thu, 27 Oct 2016, Dave Taht wrote:
>
>> I am curious if studies exists of the actual consumption in typical
>> 100mbit and above plans, vs 20Mbit, here and worldwide. Again, my
>> thesis is, aside from business (and bittorrent) users, your typical
>> 250mbit plan would have very close to the same consumption as the
>> 100mbit plan. They'd use up 250mbits for a couple hours a month, but
>> that's it.
>>
>> I am increasingly convinced that without a killer application that
>> requires it,
>> we've hit "peak bandwidth".
>
>
> You sound like my College Professor from the early 90's who said that the
> networks were now going to be so fast that there was no way that users would
> need all the available bandwidth, and that it was up to the students in the
> class to invent new uses :-)
>
> Then the web happened.

Yes, it did, and it drove adoption for a long time. Given the
interactivity required for web, however, the demand curve for more
bandwidth for it has tapered off. Web page size growth, which looked
exponential back 2012 is now linear, as both web sites "tightened up"
things, and folks installed adblockers everywhere.

I have the data on this somewhere...

> any declaration of 'peak bandwidth' is only a temporary state.

Well, even then, in the early 90s, we had a good grip on the basic requirements
for everything that is now deployed today - video streaming,
videoconferencing, etc. Web traffic is now *trivial* compared to that.

Sure - there may be a killer app in the future (AR? VR?) but it is
hard to predict that! And my overall point is that when you are on one
slope of an S-curve, it's hard to see where it will peak - or start
again.

Take, for example, the over-optimistic fiber build-out that
essentially terminated in 2000 - it's taken 16 years to use all that
up....



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org

Re: [Bloat] 22 seconds til bloat on gfiber?

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 1323 bytes --]

On Thu, Oct 27, 2016 at 12:30 PM, David Lang <david@lang.hm> wrote:

> On Thu, 27 Oct 2016, Dave Taht wrote:
>
>>
>> I am increasingly convinced that without a killer application that
>> requires it,
>> we've hit "peak bandwidth".
>>
>
> You sound like my College Professor from the early 90's who said that the
> networks were now going to be so fast that there was no way that users
> would need all the available bandwidth, and that it was up to the students
> in the class to invent new uses :-)
>
> Then the web happened.
>
> any declaration of 'peak bandwidth' is only a temporary state.


I thought 1080p was going to be close to the high-water-mark for video
bandwidth needed, and then 4K came out, and now I'm seeing 8K cameras from
RED.  Although we're deep into diminishing returns on video, I'm sure there
will be other applications...  I know that even with ~150Mbps service, I'm
routinely at that limit, and wanting more.  But I want upload more than I
want download.  12Mbps is peanuts.  I think as we get more cloud
photo/video stuff happening, there will be a push for faster upload speeds
(mostly for backups and pushing to the cloud).  100 DSLR images (24MP,
losslessly compressed) take about 30 minutes to upload at 12Mbps...

12Mbps is only 5.4GB/hour...  I'd love to have 150Mbps synchronous.

-Aaron

[-- Attachment #2: Type: text/html, Size: 1880 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 446 bytes --]

> Take, for example, the over-optimistic fiber build-out that
> essentially terminated in 2000 - it's taken 16 years to use all that
> up....
>

That sounds like it was in the right ballpark.  Trenching is so expensive,
only doing it every couple decades sounds like a reasonable plan (even
better if you trench conduit that you can run replaceable cables in (which
is what AT&T did when they took U-Verse into the Santa Cruz mountains).

-Aaron

[-- Attachment #2: Type: text/html, Size: 716 bytes --]

Re: [Bloat] 22 seconds til bloat on gfiber?

[Mikael Abrahamsson]

On Thu, 27 Oct 2016, Dave Taht wrote:

> interactive, once basic bandwidth needs are slaked, which starts to 
> happen once you crack the largest typical load (which these days is 4k 
> video streaming).
>
> gbit fiber is *way* on the unneeded side of the demand curve for home users.

I can make up credible scenarios where a home with 4-5 people would need 
200-300 megabit/s of reliably available bandwidth, plus downloading 
something large, then you can make use of a gig. So not "way" unneeded.

I do have problems coming up with scenarios where you need more than a 
gig.

> which kind of points out that you need business users to use it all up.

Business users actually use *less* bw than residential. People typically 
don't watch 4k video streams at work.

Btw, what does that report say? I don't want to spend money on it.

-- 
Mikael Abrahamsson    email: swmike@swm.pp.se

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jan Ceuleers]

On 27/10/16 21:48, Aaron Wood wrote:

> That sounds like it was in the right ballpark.  Trenching is so
> expensive, only doing it every couple decades sounds like a reasonable
> plan (even better if you trench conduit that you can run replaceable
> cables in (which is what AT&T did when they took U-Verse into the Santa
> Cruz mountains).

Yes, except that in the years leading up to 2000/2001 everyone had the
attitude "build it and they will come", meaning that there was a race
for everyone to get their own fiber into the ground which actually drove
the cost up rather than down, due to a lack of capacity in the
contracting sector, and also because lots of governments and local
authorities saw this stqggering peak in network construction activity as
an opportunity to make lots of money from rights of way.

Re: [Bloat] 22 seconds til bloat on gfiber?

[Klatsky, Carl]

Was forwarded a story today about someone using a 10Gbps residential broadband connection:

http://motherboard.vice.com/read/10-gbps-fiber-internet-fastest-home-internet-in-the-united-states

Regards,
Carl Klatsky

On 27/10/16 21:48, Aaron Wood wrote:

> That sounds like it was in the right ballpark.  Trenching is so 
> expensive, only doing it every couple decades sounds like a reasonable 
> plan (even better if you trench conduit that you can run replaceable 
> cables in (which is what AT&T did when they took U-Verse into the 
> Santa Cruz mountains).

Yes, except that in the years leading up to 2000/2001 everyone had the attitude "build it and they will come", meaning that there was a race for everyone to get their own fiber into the ground which actually drove the cost up rather than down, due to a lack of capacity in the contracting sector, and also because lots of governments and local authorities saw this stqggering peak in network construction activity as an opportunity to make lots of money from rights of way.


_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/bloat

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jonathan Morton]

> On 27 Oct, 2016, at 23:17, Mikael Abrahamsson <swmike@swm.pp.se> wrote:
> 
> I do have problems coming up with scenarios where you need more than a gig.

As long as typical consumer PCs have GigE ports and nothing (reliably) faster, more than 1Gbps at the last mile will be a tough sell.  That should be obvious.

In fact, there are few home applications where >1Gbps matters even over the LAN.  This I think will keep GigE as the typical standard for individual PCs for the foreseeable future.

That’s not to say it’s *impossible* to sell 4Gbps or 10Gbps connections.  You could do it by bundling a multi-port switch with a sufficiently fast uplink port, and sell it as “a full gigabit for each of N computers”.  The most obvious customers to target might be apartment complexes or entire villages, who could share such a connection over a large number of users and defray a relatively high installation cost.

 - Jonathan Morton

Re: [Bloat] 22 seconds til bloat on gfiber?

[Mikael Abrahamsson]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 1154 bytes --]

On Wed, 7 Dec 2016, Jonathan Morton wrote:

> That’s not to say it’s *impossible* to sell 4Gbps or 10Gbps connections. 
> You could do it by bundling a multi-port switch with a sufficiently fast 
> uplink port, and sell it as “a full gigabit for each of N computers”. 
> The most obvious customers to target might be apartment complexes or 
> entire villages, who could share such a connection over a large number 
> of users and defray a relatively high installation cost.

I believe this is what Comcast is doing for their 2 gigabit/s service, and 
why Netgear released their X10 with SFP+ uplink.

I've been told Comcast does SFP+ handoff, and this device seems to be 
tailor made for use with such a service.

Otoh it seems that 2.5GE and 5GE is going to be a thing in the not so 
distant future, I've been told 2017 will see shipping products for this 
that will be at a better price point that 10GE is currently at (which 
means quite expensive).

So I imagine we'll be seeing high end "home routers" with built in L2 
switches that have 1/2.5/5GE support to cater for this market in 2017.

-- 
Mikael Abrahamsson    email: swmike@swm.pp.se

Re: [Bloat] 22 seconds til bloat on gfiber?

[Jonathan Foulkes]

[-- Attachment #1: Type: text/plain, Size: 840 bytes --]

Dave, amen to that. Please do write that piece.

Totally agree that modern internet usage benefits most from short RTTs, low jitter and low latencies.
Heck, even 12Mbps DSL lets me enjoy most use cases with no problems.

Jonathan


> On Oct 27, 2016, at 1:58 PM, bloat-request@lists.bufferbloat.net wrote:
> 
> When you are on one side of an S curve, it's hard to see where it flattens out.
> 
> I've been meaning to research and write a piece called "have we
> reached 'Peak Bandwidth'"? for a while now.
> 
> My thesis is that what users actually want is short RTTs for
> interactive, once basic bandwidth needs are slaked, which starts to
> happen once you crack the largest typical load (which these days is 4k
> video streaming).
> 
> gbit fiber is *way* on the unneeded side of the demand curve for home users.


[-- Attachment #2: Type: text/html, Size: 7441 bytes --]