OSP multi-fiber Network-to Network Interface - Recommendations requested
lorell at hathcock.org
Tue Apr 28 05:28:05 UTC 2015
I seek information from people that have already done this kind of thing
before. Admittedly, this may be lower down the OSI model than many of them
go, but there are some lurkers among the bunch.
So, re-phrased, I would ask "What are industry standard, best practices when
designing an OSP multi-fiber NNI?"
From: Roy Hirst [mailto:rhirst at xkl.com]
Sent: Monday, April 27, 2015 11:59 AM
To: Lorell Hathcock
Subject: Re: OSP multi-fiber Network-to Network Interface - Recommendations
The silence of the experts is indeed rare. At my reading, though, it was not
very clear what you were asking the alias for. You have some perfectly
reasonable questions re negotiation, and a project to design presumably
soon. Sounds exciting. What did you want to ask, I may be able to help.
Roy Hirst | 425-556-5773 | XKL LLC | xkl.com
On 4/27/15 9:23 AM, Lorell Hathcock wrote:
It's rare that NANOG is speechless on an issue. Have I stumped the experts?
From: NANOG [mailto:nanog-bounces at nanog.org] On Behalf Of
lorell at hathcock.org <mailto:lorell at hathcock.org>
Sent: Saturday, April 25, 2015 10:53 AM
To: nanog at nanog.org <mailto:nanog at nanog.org>
Subject: FW: OSP multi-fiber Network-to Network Interface - Recommendations
The purpose of this email is to discuss information, standards,
recommendations, et cetera about interconnect solutions considering the
parameters contained herein. We believe the correct term is a multi-fiber
network to network interface.
My firm has made an IRU agreement with a municipality to use each other's
In most of the City's OSP, they have a 96 strand count fiber with eight
buffer tubes (each buffer tube having 12 fibers). They have dedicated the
black buffer tube for our use (again, 12 strands)
We have yet to build any OSP fiber plant. When we do and when we
interconnect with the City's fiber, we will extend a minimum of 96 fibers.
When our plant extends in public right of way, we will interconnect 84
strands (maybe 7 buffer tubes with 12 strands each) of our fiber to the City
and keep a minimum of one for our ourselves. It is highly likely that we
will pull a much higher fiber count cable to give ourselves additional
fibers beyond just 12 strands. On certain projects, when outside City
limits and/or on private right of way within City limits, we are required to
give them 12 strands of fiber.
When we interconnect with their fiber, we must consider the following:
1. How many strands with which will we interconnect?
a. If we are interconnecting in the middle of a City span, we must
think about interconnecting with North-bound and South-bound fibers. (12
fibers for us going in two directions and as many as 84 fibers for them
going in two directions).
b. If we are connecting at the end of a City span, we must consider
just the South-bound fibers and the interconnection between our OSP.
2. Available real estate for placing vaults, pedestals, FDHs, et
3. The likelihood of damage from accidents on the adjacent roads.
4. The likelihood of water filling up underground vaults.
5. dB loss resulting from splices, interconnects, et cetera.
6. Scalability and future growth.
7. Other considerations?
In our discussions with the City, we have contemplated a dual cabinet system
where we ask the following questions to determine how to load those
1. Where is the proposed interconnect in terms of real estate and
adjacent traffic? (Find a safe place with enough real estate).
2. How many fibers will interconnect from our network to theirs?
(either 12 or 84 for each direction we take from the site - most likely just
3. How many fibers will interconnect from their network to ours? (just
12 at most, but likely in at least two directions - because we are cutting
into their fiber mid-span)
Once those questions are answered, then we can design and build the
Also, we want to be cost effective in this design.
Thanks in advance for a push in the right direction.
Chief Technology Officer
SolStar Network, LLC
The information contained in this e-mail message may be privileged,
confidential and protected from disclosure. If you are not the intended
recipient, any dissemination, distribution or copying is strictly
prohibited. If you think that you have received this e-mail message in
error, please e-mail the sender at the above e-mail address.
More information about the NANOG