[134070] in North American Network Operators' Group
Re: inquiry on using POS
daemon@ATHENA.MIT.EDU (Mikael Abrahamsson)
Thu Dec 23 01:01:20 2010
Date: Thu, 23 Dec 2010 07:00:24 +0100 (CET)
From: Mikael Abrahamsson <swmike@swm.pp.se>
To: Jinwha Chung <chungjinwha@gmail.com>
In-Reply-To: <000401cba24b$8793d310$96bb7930$@com>
Cc: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
This message is in MIME format. The first part should be readable text,
while the remaining parts are likely unreadable without MIME-aware tools.
---137064504-1195413672-1293084024=:27193
Content-Type: TEXT/PLAIN; charset=ks_c_5601-1987; format=flowed
Content-Transfer-Encoding: 8BIT
On Thu, 23 Dec 2010, Jinwha Chung wrote:
>> From the datasheet, the other one can support point-to-point connection up
> to 80km using this DCU. Dispersion compensating unit.
>
> I¡¯ve talked about this configuration with people and no one would
> recommend those kind of things.
There is nothing saying this won't work. I'd gladly implement this (if you
by this mean the DPSK or dunobinary cards with g.709).
The advantage of this is that you get FEC and can see what margin you have
until the link starts to give post-FEC errors.
> Personally, I prefer 10G Ethernet with XENPAK ZR optics.
If dark fiber is available, this is much much cheaper, but then you have
to load balance and the customer traffic might not be possible to load
balance properly on 4x10GE, but if it is, this is definitely a viable
option.
Middle ground would be the 4 port 10GE g.709 card with FEC if you really
feel you need indications of error rate constantly. It's cheaper than the
40G card, but most likely more expensive than a 4 port 10GE card with ZR
optics.
--
Mikael Abrahamsson email: swmike@swm.pp.se
---137064504-1195413672-1293084024=:27193--
