RE: Last Mile Design

[Tony Wicks]

Certainly the devil is in the details, in New Zealand the access layer (GPON 
plus local transport) is largely regulated. Then Retail service providers buy 
the access component wholesale and add layer3, national backhaul etc. Retail 
for unlimited 1G/500M internet is about $75USD/month, for 100/50 you are 
looking at about 50USD/month. Key to this was the breakup of the incumbent into 
an access plus retail provider. This was done by allowing power (lines) 
companies in a few regions to win the access component contract.





From: NANOG  On Behalf Of Baldur Norddahl
Sent: Sunday, 10 February 2019 6:21 PM
To: nanog@nanog.org
Subject: Re: Last Mile Design



The FTTH rollout in Sweden has resulted in monopoly and the prices are high. 
Anything will work if you do not need to compete and you are getting financed 
by someone with money to spend.



On 2019-02-09 18:59 CET, Mikael Abrahamsson wrote:






---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

Re: Last Mile Design

[Baldur Norddahl]

The FTTH rollout in Sweden has resulted in monopoly and the prices are
high. Anything will work if you do not need to compete and you are getting
financed by someone with money to spend.

lør. 9. feb. 2019 22.05 skrev Thomas Bellman :

> On 2019-02-09 18:59 CET, Mikael Abrahamsson wrote:
>
> > For anyone saying it's "impossible" to do AE they're welcome here to
> > the nordic region and especially Sweden where PON is basically unheard
> > of. We have millions of AE connected households. I live in one of them.
>
> However, large parts (probably even most) of our FTTH deployments have
> been built, and are owned, by our municipalities, not private companies.
> And have had government subsidies.  (Sometimes outsourced to normal
> commercial companies, but those companies then have the municipality as
> their customer, not us end-users.)
>
> Even without the subsidies, I expect that changes what kind of long-
> term view is taken on the investment.  A purely commercial company might
> want a return on their investments in just a few years, and if the fiber
> plant needs to be replaced in its entirety ten years from now, that will
> be the problem of a different CEO. :-)  A municipality (or a company
> wholly owned by the municipality) is used to build roads and water pipes
> with expected lifetimes of 50 years, and might build their fiber plants
> expecting them to live 20 years or longer.
>
>
> /Bellman
>
>

Re: Last Mile Design

[Brandon Martin]

On 2/9/19 1:20 PM, Miles Fidelman wrote:
Though... getting power to 
the switches might be an issue, less so if you're deploying on power poles.


Just because you're on the power poles doesn't mean you can easily get 
permission or space to mount powered equipment on them let alone power 
at reasonable costs.


In areas with a commercial for-profit monopoly electric utility, just 
getting attachment space in the telecom zone at a reasonable price can 
be a big issue, and often putting stuff outside the telecom cable 
attachment zone is impossible.

--
Brandon Martin

Verizon having a bad routing day today?

[Christopher Morrow]

howdy!
I wonder if there's a lurking verizon/701 engineer on-list who may have a
few moments to reach me out of band? :) I've got what looks like busted
routing (or changes to peering/etc) causing me some headaches this last day
or so. I'd like to chat/email and see if there's a good reason for this OR
if it's just the vagaries of the tubes these days?

thanks!
-chris
(an actual verizon customer, still)

RE: Last Mile Design

[Tony Wicks]

In New Zealand we have a mostly (any town of about 20k population or more) 
nationwide FTTH rollout underway (government/private partnership) that is 
mostly based on GPON. Both Point to Point and Dark Fibre are available as well. 
The service is layer 2 QinQ delivered to the retail service providers, (1/16 
split on the GPON) while the fibre infrastructure provider is barred from 
retail service sales. GPON speeds generally delivered are 100/50, 200/200 and 
1G/500. In general the real world result of this is a network that performs 
fantastically for both retail and SMB. Larger businesses are often delivered 
over single strand dark Fibre, but in practice the 1G/500M service works 
extremely well for most situations. 10G over the PON network is about to start 
a trial phase, but the ready availability of DF significantly reduces the 
urgency of this (8x10G over cheap CWDM fibre mux's makes for a nice solution).



>
>Agreed - we generally do not recommend the use of GPON for our Enterprise 
>customers. However, in cases where a 3rd party partner discloses their use of 
>GPON to deliver our tails, we dumb down the SLA's and >technical capabilities 
>and advise the customer accordingly.
>
>Mark.


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

Re: Last Mile Design

[Mark Tinka]

On 9/Feb/19 21:12, Miles Fidelman wrote:

>  
>
> If you're marketing to business customers, or home office
> professionals, of families with multiple users that consume upstream
> bandwidth, AE gives you a lot of room for upside growth (assuming you
> provision the right kinds of fiber).

Agreed - we generally do not recommend the use of GPON for our
Enterprise customers. However, in cases where a 3rd party partner
discloses their use of GPON to deliver our tails, we dumb down the SLA's
and technical capabilities and advise the customer accordingly.

Mark.

Re: Last Mile Design

[Mark Tinka]

On 9/Feb/19 19:59, Mikael Abrahamsson wrote:

>   
>
> For anyone saying it's "impossible" to do AE they're welcome here to
> the nordic region and especially Sweden where PON is basically unheard
> of. We have millions of AE connected households. I live in one of them.

Love what Stokab did/are doing.

A prime example of how well things can be done if gubbermints and the
private sector are efficient.

Mark.

Re: Last Mile Design

[Mark Tinka]

On 9/Feb/19 18:07, Brandon Martin wrote:

>  
>
> Bingo.  You're fine as long as your access L3 gear speaks MPLS.  That
> does somewhat bump you out of the realm of "cheap L3 switch", but
> there are still options.

IP-capable switches that have little to no MPLS support would certainly
be cheaper than one that does.

But given the benefits of an MPLS-based Metro-E network vs. the
traditional architecture, I find current prices somewhat reasonable.

Mark.

Re: Last Mile Design

[Miles Fidelman]

On 2/9/19 4:04 PM, Thomas Bellman wrote:


On 2019-02-09 18:59 CET, Mikael Abrahamsson wrote:


For anyone saying it's "impossible" to do AE they're welcome here to
the nordic region and especially Sweden where PON is basically unheard
of. We have millions of AE connected households. I live in one of them.

However, large parts (probably even most) of our FTTH deployments have
been built, and are owned, by our municipalities, not private companies.
And have had government subsidies.  (Sometimes outsourced to normal
commercial companies, but those companies then have the municipality as
their customer, not us end-users.)

Even without the subsidies, I expect that changes what kind of long-
term view is taken on the investment.  A purely commercial company might
want a return on their investments in just a few years, and if the fiber
plant needs to be replaced in its entirety ten years from now, that will
be the problem of a different CEO. :-)  A municipality (or a company
wholly owned by the municipality) is used to build roads and water pipes
with expected lifetimes of 50 years, and might build their fiber plants
expecting them to live 20 years or longer.


Yes indeed, longer time horizon, but generally not so much subsidized as:

- having a big initial customer (municipal electric utility, water 
utility, the city or county) - a lot of municipal builds are essentially 
done for internal purposes, with service to the public as a bonus


- still, usually funded by bonds - long-term money, low rates - and 
maybe some money from the NTIA (also available to rural coops)


- a view of networks as infrastructure, with cost-recovery pricing, 
rather than as a revenue stream to milk (same as internal networks at a 
university or corporation)


By and large, there's a pretty good argument that we SHOULD be viewing 
broadband networks as infrastructure, with ownership & management to 
match.  (Fair disclosure, I used to promote that view as director of a 
non-profit policy shop, and as a consultant to municipal governments.  
I've also helped design & build big networks, for big customers - in my 
days at BBN - so it's an informed opinion :-).


Miles Fidelman


--
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra

Re: Last Mile Design

[Grant Taylor via NANOG]

On 2/9/19 1:13 PM, Baldur Norddahl wrote:

GPON is 2.4 Gbps downstream and 1.2 Gbps upstream.


Okay.

I guess I've not thought about the fact that the GPON itself might be ~> 
is asymmetric.  From my naive point of view, I see a 1G/1G symmetric 
Ethernet hand off from the ONT to my equipment.  Hence my uninformed 
understanding.


Residential users are download heavy and more than 1:2. However there is 
a big difference between average, peak and micro burst. The conclusion 
is not simple.


ACK

We typically have 60+ users on each port. We sell 1000/1000 internet. 
And yet we only get good ratings for the speed.


I've learned that people are quick to judge harshly and slow to 
complement.  Or that good or better speeds ratings are lost to other 
things like price and / or other services offered, like native IPv6 or not.


I find that many, that are sceptical about the shared bandwidth of GPON, 
forget that a typical POP might only be fed by a 10 Gbps uplink. Usually 
this has much lower bandwidth per user than the GPON link.


I remember having these discussions in the early 2000's about ADSL vs 
Cable Modem.  I wonder if some of the earlier horror stories are 
unconsciously biasing people's opinions.  Or if people quite literally 
only look at / think about the directly attached network segment.




--
Grant. . . .
unix || die



smime.p7s
Description: S/MIME Cryptographic Signature

Re: Last Mile Design

[Thomas Bellman]

On 2019-02-09 18:59 CET, Mikael Abrahamsson wrote:

> For anyone saying it's "impossible" to do AE they're welcome here to
> the nordic region and especially Sweden where PON is basically unheard
> of. We have millions of AE connected households. I live in one of them.

However, large parts (probably even most) of our FTTH deployments have
been built, and are owned, by our municipalities, not private companies.
And have had government subsidies.  (Sometimes outsourced to normal
commercial companies, but those companies then have the municipality as
their customer, not us end-users.)

Even without the subsidies, I expect that changes what kind of long-
term view is taken on the investment.  A purely commercial company might
want a return on their investments in just a few years, and if the fiber
plant needs to be replaced in its entirety ten years from now, that will
be the problem of a different CEO. :-)  A municipality (or a company
wholly owned by the municipality) is used to build roads and water pipes
with expected lifetimes of 50 years, and might build their fiber plants
expecting them to live 20 years or longer.


/Bellman



signature.asc
Description: OpenPGP digital signature

Re: Last Mile Design

[Miles Fidelman]

There is that.


On 2/9/19 3:27 PM, Mike Hammett wrote:
The biggest use of bandwidth as the IoT buzzword comes to fruition is 
exploits.




-
Mike Hammett
Intelligent Computing Solutions 

Midwest Internet Exchange 

The Brothers WISP 


*From: *"Miles Fidelman" 
*To: *"Mike Hammett" 
*Cc: *nanog@nanog.org
*Sent: *Saturday, February 9, 2019 2:26:13 PM
*Subject: *Re: Last Mile Design

I expect things are going to change as IoT takes off - security 
cameras, baby monitors, start to push video upstream - that makes a 
difference.



And then there are the efforts of cell carriers to push traffic onto 
home wifi - more and more facetime video will also add load.



Miles


On 2/9/19 3:14 PM, Mike Hammett wrote:

Electrical consumption of the equipment is different and then the
environmental conditioning that larger electronic load.

Let's not forget that actual consumer bit consumption changes very
little whether they have 20 megs or 2 gigs provisioned and available.



-
Mike Hammett
Intelligent Computing Solutions 


Midwest Internet Exchange 


The Brothers WISP 



*From: *"Miles Fidelman" 
*To: *nanog@nanog.org
*Sent: *Saturday, February 9, 2019 12:20:36 PM
*Subject: *Re: Last Mile Design

Speaking of which, the Grant County Public Utility District
(Washington
State), has wired active ethernet all over their rural county.

Seems to me that the cost difference between splitters & switches
is a
pretty minor component of deploying FTTH - the costs are in the
trenching, and the fiber.  What you put on the poles, or in the lawn
furniture, is a pretty minor cost component. Though... getting
power to
the switches might be an issue, less so if you're deploying on
power poles.

Miles Fidelman

On 2/9/19 12:59 PM, Mikael Abrahamsson wrote:
> On Sat, 9 Feb 2019, Mark Tinka wrote:
>
>> If I had to build a consumer broadband network and had the budget
>> (and owned the fibre) to do so, I'd definitely always choose
Active-E:
>
> For anyone saying it's "impossible" to do AE they're welcome
here to
> the nordic region and especially Sweden where PON is basically
unheard
> of. We have millions of AE connected households. I live in one
of them.
>
-- 
In theory, there is no difference between theory and practice.

In practice, there is.   Yogi Berra


--
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra


--
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra

Re: Last Mile Design

[Mike Hammett]

The biggest use of bandwidth as the IoT buzzword comes to fruition is exploits. 




- 
Mike Hammett 
Intelligent Computing Solutions 

Midwest Internet Exchange 

The Brothers WISP 

- Original Message -

From: "Miles Fidelman"  
To: "Mike Hammett"  
Cc: nanog@nanog.org 
Sent: Saturday, February 9, 2019 2:26:13 PM 
Subject: Re: Last Mile Design 


I expect things are going to change as IoT takes off - security cameras, baby 
monitors, start to push video upstream - that makes a difference. 


And then there are the efforts of cell carriers to push traffic onto home wifi 
- more and more facetime video will also add load. 


Miles 



On 2/9/19 3:14 PM, Mike Hammett wrote: 



Electrical consumption of the equipment is different and then the environmental 
conditioning that larger electronic load. 

Let's not forget that actual consumer bit consumption changes very little 
whether they have 20 megs or 2 gigs provisioned and available. 




- 
Mike Hammett 
Intelligent Computing Solutions 

Midwest Internet Exchange 

The Brothers WISP 

- Original Message -

From: "Miles Fidelman"  
To: nanog@nanog.org 
Sent: Saturday, February 9, 2019 12:20:36 PM 
Subject: Re: Last Mile Design 

Speaking of which, the Grant County Public Utility District (Washington 
State), has wired active ethernet all over their rural county. 

Seems to me that the cost difference between splitters & switches is a 
pretty minor component of deploying FTTH - the costs are in the 
trenching, and the fiber. What you put on the poles, or in the lawn 
furniture, is a pretty minor cost component. Though... getting power to 
the switches might be an issue, less so if you're deploying on power poles. 

Miles Fidelman 

On 2/9/19 12:59 PM, Mikael Abrahamsson wrote: 
> On Sat, 9 Feb 2019, Mark Tinka wrote: 
> 
>> If I had to build a consumer broadband network and had the budget 
>> (and owned the fibre) to do so, I'd definitely always choose Active-E: 
> 
> For anyone saying it's "impossible" to do AE they're welcome here to 
> the nordic region and especially Sweden where PON is basically unheard 
> of. We have millions of AE connected households. I live in one of them. 
> 
-- 
In theory, there is no difference between theory and practice. 
In practice, there is.  Yogi Berra 




-- 
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra 

Re: Last Mile Design

[Miles Fidelman]

I expect things are going to change as IoT takes off - security cameras, 
baby monitors, start to push video upstream - that makes a difference.



And then there are the efforts of cell carriers to push traffic onto 
home wifi - more and more facetime video will also add load.



Miles


On 2/9/19 3:14 PM, Mike Hammett wrote:
Electrical consumption of the equipment is different and then the 
environmental conditioning that larger electronic load.


Let's not forget that actual consumer bit consumption changes very 
little whether they have 20 megs or 2 gigs provisioned and available.




-
Mike Hammett
Intelligent Computing Solutions 

Midwest Internet Exchange 

The Brothers WISP 


*From: *"Miles Fidelman" 
*To: *nanog@nanog.org
*Sent: *Saturday, February 9, 2019 12:20:36 PM
*Subject: *Re: Last Mile Design

Speaking of which, the Grant County Public Utility District (Washington
State), has wired active ethernet all over their rural county.

Seems to me that the cost difference between splitters & switches is a
pretty minor component of deploying FTTH - the costs are in the
trenching, and the fiber.  What you put on the poles, or in the lawn
furniture, is a pretty minor cost component. Though... getting power to
the switches might be an issue, less so if you're deploying on power 
poles.


Miles Fidelman

On 2/9/19 12:59 PM, Mikael Abrahamsson wrote:
> On Sat, 9 Feb 2019, Mark Tinka wrote:
>
>> If I had to build a consumer broadband network and had the budget
>> (and owned the fibre) to do so, I'd definitely always choose Active-E:
>
> For anyone saying it's "impossible" to do AE they're welcome here to
> the nordic region and especially Sweden where PON is basically unheard
> of. We have millions of AE connected households. I live in one of them.
>
--
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra



--
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra

Re: Last Mile Design

[Miles Fidelman]

On 2/9/19 2:51 PM, Grant Taylor via NANOG wrote:


On 2/9/19 12:12 PM, Miles Fidelman wrote:
With early PON designs, upstream bandwidth was horrible. Not 
particularly useful if you're doing things like remote backup, or 
video chatting, or running a server (business grade service). GPON 
does better on upstream bandwidth, but it's still asymmetric.


Intriguing.

I would have not considered my municipal GPON to be asymmetric. Well, 
not as such.  Routinely, when I do speed tests I get better upstream 
speeds than I do downstream speeds.  (More below.)


If you're marketing to business customers, or home office 
professionals, of families with multiple users that consume upstream 
bandwidth, AE gives you a lot of room for upside growth (assuming you 
provision the right kinds of fiber).


Are you referring to the dedicated bandwidth between the CPU and the 
AE equipment?  Or the fact that bandwidth feeding the GPON and all 
subscribers is aggregate?



I'm thinking about the backside.  Generally there's a lot more 
downstream bandwidth to distribute, and not a lot of upstream 
bandwidth.  Makes a lot of sense if you're a content provider & expect 
your customers to be passive consumers (also, considering that a lot of 
that bandwidth might be used for things other than IP packets).




I have attributed the asymmetry in my speed tests to be that most 
people on my GPON are predominantly downloading, thus consuming 
aggregate download bandwidth.  Conversely, few are uploading more than 
requests, thus using relatively little of the aggregate upload bandwidth.


Probably the case.  But if you're in an area with a lot of home office 
users, or gamers, or business grade customers running servers, your 
experience might be different.




Do I see asymmetry?  Yes.  Is it truly asymmetric?  I don't think so.  
I think is just based on consumption of aggregate bandwidth.


I have no idea if this is normal for GPON or not.  Hence one of the 
reasons that I'm finding this thread enlightening.



The SPECS are asymmetric, as is the technology when you take into 
account allocation of bandwidth between downstream video & IP services.


Miles




--
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra

Re: Last Mile Design

[Mike Hammett]

Electrical consumption of the equipment is different and then the environmental 
conditioning that larger electronic load. 

Let's not forget that actual consumer bit consumption changes very little 
whether they have 20 megs or 2 gigs provisioned and available. 




- 
Mike Hammett 
Intelligent Computing Solutions 

Midwest Internet Exchange 

The Brothers WISP 

- Original Message -

From: "Miles Fidelman"  
To: nanog@nanog.org 
Sent: Saturday, February 9, 2019 12:20:36 PM 
Subject: Re: Last Mile Design 

Speaking of which, the Grant County Public Utility District (Washington 
State), has wired active ethernet all over their rural county. 

Seems to me that the cost difference between splitters & switches is a 
pretty minor component of deploying FTTH - the costs are in the 
trenching, and the fiber. What you put on the poles, or in the lawn 
furniture, is a pretty minor cost component. Though... getting power to 
the switches might be an issue, less so if you're deploying on power poles. 

Miles Fidelman 

On 2/9/19 12:59 PM, Mikael Abrahamsson wrote: 
> On Sat, 9 Feb 2019, Mark Tinka wrote: 
> 
>> If I had to build a consumer broadband network and had the budget 
>> (and owned the fibre) to do so, I'd definitely always choose Active-E: 
> 
> For anyone saying it's "impossible" to do AE they're welcome here to 
> the nordic region and especially Sweden where PON is basically unheard 
> of. We have millions of AE connected households. I live in one of them. 
> 
-- 
In theory, there is no difference between theory and practice. 
In practice, there is.  Yogi Berra 

Re: Last Mile Design

[Baldur Norddahl]

GPON is 2.4 Gbps downstream and 1.2 Gbps upstream. Residential users are
download heavy and more than 1:2. However there is a big difference between
average, peak and micro burst. The conclusion is not simple.

We typically have 60+ users on each port. We sell 1000/1000 internet. And
yet we only get good ratings for the speed.

I find that many, that are sceptical about the shared bandwidth of GPON,
forget that a typical POP might only be fed by a 10 Gbps uplink. Usually
this has much lower bandwidth per user than the GPON link.

Regards

Baldur


lør. 9. feb. 2019 20.52 skrev Grant Taylor via NANOG :

> On 2/9/19 12:12 PM, Miles Fidelman wrote:
> > With early PON designs, upstream bandwidth was horrible. Not
> > particularly useful if you're doing things like remote backup, or video
> > chatting, or running a server (business grade service). GPON does better
> > on upstream bandwidth, but it's still asymmetric.
>
> Intriguing.
>
> I would have not considered my municipal GPON to be asymmetric.  Well,
> not as such.  Routinely, when I do speed tests I get better upstream
> speeds than I do downstream speeds.  (More below.)
>
> > If you're marketing to business customers, or home office professionals,
> > of families with multiple users that consume upstream bandwidth, AE
> > gives you a lot of room for upside growth (assuming you provision the
> > right kinds of fiber).
>
> Are you referring to the dedicated bandwidth between the CPU and the AE
> equipment?  Or the fact that bandwidth feeding the GPON and all
> subscribers is aggregate?
>
> I have attributed the asymmetry in my speed tests to be that most people
> on my GPON are predominantly downloading, thus consuming aggregate
> download bandwidth.  Conversely, few are uploading more than requests,
> thus using relatively little of the aggregate upload bandwidth.
>
> Do I see asymmetry?  Yes.  Is it truly asymmetric?  I don't think so.  I
> think is just based on consumption of aggregate bandwidth.
>
> I have no idea if this is normal for GPON or not.  Hence one of the
> reasons that I'm finding this thread enlightening.
>
>
>
> --
> Grant. . . .
> unix || die
>
>

Re: Last Mile Design

[Grant Taylor via NANOG]

On 2/9/19 12:12 PM, Miles Fidelman wrote:
With early PON designs, upstream bandwidth was horrible. Not 
particularly useful if you're doing things like remote backup, or video 
chatting, or running a server (business grade service). GPON does better 
on upstream bandwidth, but it's still asymmetric.


Intriguing.

I would have not considered my municipal GPON to be asymmetric.  Well, 
not as such.  Routinely, when I do speed tests I get better upstream 
speeds than I do downstream speeds.  (More below.)


If you're marketing to business customers, or home office professionals, 
of families with multiple users that consume upstream bandwidth, AE 
gives you a lot of room for upside growth (assuming you provision the 
right kinds of fiber).


Are you referring to the dedicated bandwidth between the CPU and the AE 
equipment?  Or the fact that bandwidth feeding the GPON and all 
subscribers is aggregate?


I have attributed the asymmetry in my speed tests to be that most people 
on my GPON are predominantly downloading, thus consuming aggregate 
download bandwidth.  Conversely, few are uploading more than requests, 
thus using relatively little of the aggregate upload bandwidth.


Do I see asymmetry?  Yes.  Is it truly asymmetric?  I don't think so.  I 
think is just based on consumption of aggregate bandwidth.


I have no idea if this is normal for GPON or not.  Hence one of the 
reasons that I'm finding this thread enlightening.




--
Grant. . . .
unix || die



smime.p7s
Description: S/MIME Cryptographic Signature

Re: Last Mile Design

[Miles Fidelman]

On 2/9/19 1:44 PM, Grant Taylor via NANOG wrote:



So let me ask this:  Are there any functional reasons to favor AE over 
PON /within/ the lifecycle of a deployment?  Does one methodology 
offer any significant advantages or disadvantages over the other?  If 
so, is (are) the pro(s) / con(s) applicable to specific use case(s)?


With early PON designs, upstream bandwidth was horrible. Not 
particularly useful if you're doing things like remote backup, or video 
chatting, or running a server (business grade service). GPON does better 
on upstream bandwidth, but it's still asymmetric.


If you're marketing to business customers, or home office professionals, 
of families with multiple users that consume upstream bandwidth, AE 
gives you a lot of room for upside growth (assuming you provision the 
right kinds of fiber).



---

In theory, there is no difference between theory and practice. In 
practice, there is.  Yogi Berra

Re: Last Mile Design

[Grant Taylor via NANOG]

On 2/9/19 11:22 AM, Sander Steffann wrote:
Same for me. I like the architecture where the PON splitters are in 
powered roadside cabinets (even though the splitter is passive). That 
way the ISP can convert it to AE at any time they want. The architectures 
where PON has been hardcoded into the design has always felt like a huge 
risk regarding future developments.


I agree that PON with splitters where you can't put Active Ethernet 
equipment is largely equivalent to solution lock-in.  But is that in and 
of itself a bad thing?  Especially when viewed in within the lifecycle 
of the network?


From an outsider n00b point of view, the things that I'm reading it 
seems that people don't like about PON are largely it's inflexibility to 
be able to be converted to Active Ethernet without careful forethought 
and planning at construction time of the fiber network to allow it to 
change in the future.


In some ways, I've heard of the industry having this, or a very similar 
discussion for 25 years.  Twisted pair (Cat 3 vs Cat 5 vs Cat 5e vs Cat 
6) vs coax (RG 59 vs RG 6 vs F-11) vs fiber (OS1 vs OS2 vs OM1 vs OM2 vs 
OM3 vs OM4 vs OM5) vs RF (myriad of options).  Included to topology 
designs equating to lock-in.


However, none of this seems to be related to how functional any given 
design is.  I guess perhaps that the subject "Last Mile Design" does 
encourage discussions about topology and technology.


So let me ask this:  Are there any functional reasons to favor AE over 
PON /within/ the lifecycle of a deployment?  Does one methodology offer 
any significant advantages or disadvantages over the other?  If so, is 
(are) the pro(s) / con(s) applicable to specific use case(s)?


Remember that there are LOTs of ways to do things.  I'm trying to glean 
what makes one method better or worse than another, possibly for 
different types of deployments.





--
Grant. . . .
unix || die



smime.p7s
Description: S/MIME Cryptographic Signature

Re: Last Mile Design

[Baldur Norddahl]

 It is not impossible just more expensive. Incidentally here in Denmark we
have TDC now converting active ethernet to GPON.

lør. 9. feb. 2019 19.01 skrev Mikael Abrahamsson :

> On Sat, 9 Feb 2019, Mark Tinka wrote:
>
> > If I had to build a consumer broadband network and had the budget (and
> > owned the fibre) to do so, I'd definitely always choose Active-E:
>
> For anyone saying it's "impossible" to do AE they're welcome here to the
> nordic region and especially Sweden where PON is basically unheard of. We
> have millions of AE connected households. I live in one of them.
>
> --
> Mikael Abrahamssonemail: swm...@swm.pp.se
>

Re: Last Mile Design

[Clayton Zekelman]

We have around 60,000 homes passed with GPON architecture.  I'm not 
really sure how we would have built that with active roadside 
cabinets, and still have been able to maintain any sort cost control.


If we did it with a home run individual fibre scheme hauled back to a 
central POP, the frame would have been massive and the power and 
cooling requirements would have made the entire project unfeasible.


Maybe the economics are different in other markets.

Because PON is so widely deployed, you can count on vendors coming up 
with capacity increases (NG, X, etc.) to support the installed base 
of infrastructure. Verizon alone will drive that market.


 From a purist point of view, AE is a nice idea, but it really isn't 
necessary for now or the foreseeable future.



At 01:22 PM 09/02/2019, Sander Steffann wrote:

Hi Mark,

> My preference, for the home, would be Active-E. But I do 
understand the economics that may support PON, and my position on 
that has softened over the years.


Same for me. I like the architecture where the PON splitters are in 
powered roadside cabinets (even though the splitter is passive). 
That way the ISP can convert it to AE at any time they want. The 
architectures where PON has been hardcoded into the design has 
always felt like a huge risk regarding future developments.


Cheers,
Sander




--

Clayton Zekelman
Managed Network Systems Inc. (MNSi)
3363 Tecumseh Rd. E
Windsor, Ontario
N8W 1H4

tel. 519-985-8410
fax. 519-985-8409

Re: Last Mile Design

[Sander Steffann]

Hi Mark,

> My preference, for the home, would be Active-E. But I do understand the 
> economics that may support PON, and my position on that has softened over the 
> years.

Same for me. I like the architecture where the PON splitters are in powered 
roadside cabinets (even though the splitter is passive). That way the ISP can 
convert it to AE at any time they want. The architectures where PON has been 
hardcoded into the design has always felt like a huge risk regarding future 
developments.

Cheers,
Sander



signature.asc
Description: Message signed with OpenPGP

Re: Last Mile Design

[Miles Fidelman]

Speaking of which, the Grant County Public Utility District (Washington 
State), has wired active ethernet all over their rural county.


Seems to me that the cost difference between splitters & switches is a 
pretty minor component of deploying FTTH - the costs are in the 
trenching, and the fiber.  What you put on the poles, or in the lawn 
furniture, is a pretty minor cost component. Though... getting power to 
the switches might be an issue, less so if you're deploying on power poles.


Miles Fidelman

On 2/9/19 12:59 PM, Mikael Abrahamsson wrote:

On Sat, 9 Feb 2019, Mark Tinka wrote:

If I had to build a consumer broadband network and had the budget 
(and owned the fibre) to do so, I'd definitely always choose Active-E:


For anyone saying it's "impossible" to do AE they're welcome here to 
the nordic region and especially Sweden where PON is basically unheard 
of. We have millions of AE connected households. I live in one of them.



--
In theory, there is no difference between theory and practice.
In practice, there is.   Yogi Berra

Re: Last Mile Design

[Mikael Abrahamsson]

On Sat, 9 Feb 2019, Mark Tinka wrote:

If I had to build a consumer broadband network and had the budget (and 
owned the fibre) to do so, I'd definitely always choose Active-E:   


For anyone saying it's "impossible" to do AE they're welcome here to the 
nordic region and especially Sweden where PON is basically unheard of. We 
have millions of AE connected households. I live in one of them.


--
Mikael Abrahamssonemail: swm...@swm.pp.se

Re: Last Mile Design

[Brandon Martin]

On 2/9/19 2:13 AM, Mark Tinka wrote:

My assumption is that you'd be running full IP/MPLS all the way into the
Access. In that case, what I'm saying is that you can run EoMPLS to
deliver the service.


Bingo.  You're fine as long as your access L3 gear speaks MPLS.  That 
does somewhat bump you out of the realm of "cheap L3 switch", but there 
are still options.

--
Brandon Martin

Re: Last Mile Design

[Mark Tinka]

My preference, for the home, would be Active-E. But I do understand the
economics that may support PON, and my position on that has softened
over the years. My service provider delivers their FTTH service to me
via PON, and for the most part, it's been all good.

That said, I was particularly impressed with what CDE Lightband did in
Clarksville, Tennessee, where they deployed their FTTH network with
Active-E using Brocade to over 60,000 subscribers:

    https://www.youtube.com/watch?v=cV1nYGl_Bjc

If I had to build a consumer broadband network and had the budget (and
owned the fibre) to do so, I'd definitely always choose Active-E:   

In South Africa, we have an access network operator that uses Active-E
primarily to deliver their service, making it perhaps the only FTTH
provider not using PON to do this. I find this quite fascinating.

Mark.

On 9/Feb/19 12:59, Ben Cannon wrote:
> I should probably have mentioned that in this sense I view “urban” as
> exclusive to “single family homes” - meaning I’m talking about high
> density modern urban with under grounding requirements - and high rise
> residential towers.
>
> We are the opposite, we are presently enterprise, midsize, and
> exotic-small business only, and have no residential arm or support
> structure (or SLA expectations, or standards or lack thereof) of a
> residential connection.
>
> -Ben.
>
> -Ben Cannon
> CEO 6x7 Networks & 6x7 Telecom, LLC 
> b...@6by7.net 
>
>
>
>> On Feb 9, 2019, at 2:54 AM, Baldur Norddahl
>> mailto:baldur.nordd...@gmail.com>> wrote:
>>
>> PON in urban areas absolutely makes sense. Maybe less in a high rise
>> area, where each building can have a small building wide network of
>> its own. But it in areas with single family homes PON is king.
>>
>> Our POPs can have up to 10 000 customers each. All on a single 96
>> fiber strand cable leading into the POP building. We have extra
>> ducts, but nothing that would allow us to change that to a point to
>> point network. That would require 100x that 96 fiber cable.
>>
>> With extra ducts it would be possible to rebuild from PON to point to
>> point. But it would require massive investments. Basically you would
>> have to invest all that we saved by building PON. For starters, you
>> would have to have many more POPs.
>>
>> And yes, there are splitters in the hand holes. This is not what
>> stops you from rebuilding from PON. It is the fact that we never paid
>> for extra fiber. The backbone in a sub area is typically build with a
>> 24 fiber strand cable. Because fibers are not free and are actually
>> quite expensive as the number of fibers grow and the distances get
>> longer. We can do a few point to point connections, for example if we
>> need to deliver a commercial service or for our own needs (to connect
>> POPs etc).
>>
>> We are not big on commercial services. But if we were, I would use
>> WDM splitters for that. Or the long awaited 10G PON if that ever
>> arrives and turns out at a price point that works.
>>
>> Regards,
>>
>> Baldur
>>  
>

Re: Comcast - NTT seeing congestion in Chicago at 350 Cermak

[Job Snijders]

Hi,

I'll follow up off list.

Kind regards,

Job

On Sat, Feb 09, 2019 at 03:05:22AM +, Erik Sundberg wrote:
> Comcast\NTT,
> 
> I am seeing a bit of congestion between the NTT and Comcast connection in 
> Chicago. Can you guys take a look at this?
> 
> 
> Normally this is a sub 10ms path, it running at 100ms.
> 
> 
> 
> speedtest (0.0.0.0)   Fri 
> Feb  8 20:23:49 2019
> Keys:  Help   Display mode   Restart statistics   Order of fields   quit
>   Packets 
>   Pings
>  Host   Loss%   Snt   Last   
> Avg  Best  Wrst StDev
>  1. 45.61.24.33  0.0%   8621.0   
> 1.1   0.9  24.9   1.6
>  2. te-0-0-25.ear2.chi2.us.nitelusa.net  0.0%   8620.9   
> 0.9   0.8  54.4   2.2
>  3. te-0-0-25.ear1.chi2.us.nitelusa.net  0.0%   8621.5   
> 1.2   0.9  34.5   1.6
>  4. te-0-0-24.ear1.chi1.us.nitelusa.net  0.0%   8621.1   
> 1.1   0.9  74.4   3.0
>  5. te-0-0-1-0.cr1.chi1.us.nitelusa.net  0.0%   8620.8   
> 0.7   0.7  13.7   0.6
>  6. xe-0-0-8-0.a02.chcgil09.us.bb.gin.ntt.net0.0%   861   42.5   
> 2.7   0.3  54.7   6.4
>  7. ae-0.comcast.chcgil09.us.bb.gin.ntt.net  0.5%   861  102.1  
> 99.1  42.0 120.2  10.3
>  8. be-10577-cr02.350ecermak.il.ibone.comcast.net0.6%   861  113.8 
> 100.7  41.6 161.0  10.7
>  9. be-7922-ar01.area4.il.chicago.comcast.net1.2%   861  107.8 
> 100.7  45.1 127.6  10.8
> 10. be-123-rur02.homewood.il.chicago.comcast.net 0.6%   861  106.9 
> 102.0  42.1 123.6  10.8
> 11. 68.87.235.2060.9%   861  102.0 
> 101.8  44.7 140.1  10.7
> 12. c-xxx.hsd1.il.comcast.net0.7%   861  103.1 
> 110.7  49.9 136.4  10.5
> 
> 
> 
> 
> CONFIDENTIALITY NOTICE: This e-mail transmission, and any documents, files or 
> previous e-mail messages attached to it may contain confidential information 
> that is legally privileged. If you are not the intended recipient, or a 
> person responsible for delivering it to the intended recipient, you are 
> hereby notified that any disclosure, copying, distribution or use of any of 
> the information contained in or attached to this transmission is STRICTLY 
> PROHIBITED. If you have received this transmission in error please notify the 
> sender immediately by replying to this e-mail. You must destroy the original 
> transmission and its attachments without reading or saving in any manner. 
> Thank you.

Re: Last Mile Design

[Ben Cannon]

I should probably have mentioned that in this sense I view “urban” as exclusive 
to “single family homes” - meaning I’m talking about high density modern urban 
with under grounding requirements - and high rise residential towers.

We are the opposite, we are presently enterprise, midsize, and exotic-small 
business only, and have no residential arm or support structure (or SLA 
expectations, or standards or lack thereof) of a residential connection.

-Ben.

-Ben Cannon
CEO 6x7 Networks & 6x7 Telecom, LLC 
b...@6by7.net 




> On Feb 9, 2019, at 2:54 AM, Baldur Norddahl  wrote:
> 
> PON in urban areas absolutely makes sense. Maybe less in a high rise area, 
> where each building can have a small building wide network of its own. But it 
> in areas with single family homes PON is king.
> 
> Our POPs can have up to 10 000 customers each. All on a single 96 fiber 
> strand cable leading into the POP building. We have extra ducts, but nothing 
> that would allow us to change that to a point to point network. That would 
> require 100x that 96 fiber cable.
> 
> With extra ducts it would be possible to rebuild from PON to point to point. 
> But it would require massive investments. Basically you would have to invest 
> all that we saved by building PON. For starters, you would have to have many 
> more POPs.
> 
> And yes, there are splitters in the hand holes. This is not what stops you 
> from rebuilding from PON. It is the fact that we never paid for extra fiber. 
> The backbone in a sub area is typically build with a 24 fiber strand cable. 
> Because fibers are not free and are actually quite expensive as the number of 
> fibers grow and the distances get longer. We can do a few point to point 
> connections, for example if we need to deliver a commercial service or for 
> our own needs (to connect POPs etc).
> 
> We are not big on commercial services. But if we were, I would use WDM 
> splitters for that. Or the long awaited 10G PON if that ever arrives and 
> turns out at a price point that works.
> 
> Regards,
> 
> Baldur
>  

Re: Last Mile Design

[Baldur Norddahl]

PON in urban areas absolutely makes sense. Maybe less in a high rise area,
where each building can have a small building wide network of its own. But
it in areas with single family homes PON is king.

Our POPs can have up to 10 000 customers each. All on a single 96 fiber
strand cable leading into the POP building. We have extra ducts, but
nothing that would allow us to change that to a point to point network.
That would require 100x that 96 fiber cable.

With extra ducts it would be possible to rebuild from PON to point to
point. But it would require massive investments. Basically you would have
to invest all that we saved by building PON. For starters, you would have
to have many more POPs.

And yes, there are splitters in the hand holes. This is not what stops you
from rebuilding from PON. It is the fact that we never paid for extra
fiber. The backbone in a sub area is typically build with a 24 fiber strand
cable. Because fibers are not free and are actually quite expensive as the
number of fibers grow and the distances get longer. We can do a few point
to point connections, for example if we need to deliver a commercial
service or for our own needs (to connect POPs etc).

We are not big on commercial services. But if we were, I would use WDM
splitters for that. Or the long awaited 10G PON if that ever arrives and
turns out at a price point that works.

Regards,

Baldur