[177889] in North American Network Operators' Group
Re: Low cost WDM gear
daemon@ATHENA.MIT.EDU (Kenneth McRae)
Sat Feb 7 15:53:27 2015
X-Original-To: nanog@nanog.org
To: Faisal Imtiaz <faisal@snappytelecom.net>
From: Kenneth McRae <kenneth.mcrae@me.com>
Date: Sat, 07 Feb 2015 20:49:16 +0000 (GMT)
Cc: NANOG <nanog@nanog.org>
Errors-To: nanog-bounces@nanog.org
That's why I engage the engineering resources on their end to make sure th=
e chosen candidate will support the use case. =C2=A0I have now performed a=
n A/B comparison and the FiberStore gear is inferior. =C2=A0Excessive loss=
on the mux and optics.=C2=A0=0A=0AOn Feb 07, 2015, at 12:44 PM, Faisal Im=
tiaz <faisal@snappytelecom.net> wrote:=0A=0AIf you pay close attention to =
the Spec Sheets, on power output, insertion loss, sensitivity, and do the =
proper calculation for your link, then using anyone's products, passive or=
active will work unless the devices do not meet specified specs.=0A=0AIf =
you don't do your homework, cals on the design, loss, and just buy stuff b=
ased on whatever, then it does not matter who the mfg. is, you are very ve=
ry likely to be surprised in a bad way.=0A=0A:)=0A=0AFaisal Imtiaz=0ASnapp=
y Internet & Telecom=0A=0A----- Original Message -----=0AFrom: "Rodrigo 1t=
elecom" <rodrigo@1telecom.com.br>=0ATo: "Kenneth McRae" <kenneth.mcrae@me.=
com>=0ACc: "NANOG" <nanog@nanog.org>=0ASent: Saturday, February 7, 2015 3:=
24:43 PM=0ASubject: Re: Low cost WDM gear=0AWhat others vendors do you usi=
ng? Here in Brazil only PADTEC have this=0Apassive solution... Some days a=
go Digitel contact me to show your multiplex=0Asolution... Is a active sol=
ution...=0AWe import this from fiberstore, but i don't know others vendors=
to buy 10G=0Asfp+ cwdm and this mux/demux...=0AEnviado via iPhone =EF=A3=BF=
=0AGrupo Connectoway=0A> Em 07/02/2015, =C3=A0s 16:04, Kenneth McRae <kenn=
eth.mcrae@me.com> escreveu:=0A>=0A> Hi Enviado,=0A>=0A> I cannot recommend=
FiberStore as I had a bad experience with them. I=0A> needed to cover onl=
y 3km from A to B side. When using 10km optics, I saw=0A> a loss of over 5=
db with their passive mux inserted into the path which=0A> created a total=
loss of over -20db which is outside of the tolerances for=0A> our equipme=
nt with 10km SFP+. Using another vendors low insertion loss=0A> mux correc=
ted our issue. I am sure if you are using an 80km optic, you=0A> may be ab=
le to tolerate a higher insertion loss to cover < 60km. I also=0A> notice =
that their CDWM optics averaged about 3db less in power output when=0A> co=
mpared to other vendors.=0A>=0A> Thanks=0A>=0A> Kenneth=0A>=0A>> On Feb 07=
, 2015, at 10:33 AM, Rodrigo 1telecom <rodrigo@1telecom.com.br>=0A>> wrote=
:=0A>>=0A>=0A>> Hi kenneth... which the distance do you have from side A t=
o side B when=0A>> you using passive solutions from fiberstore( mux and de=
mux)?=0A>> I buy this mux and demux(4 channels single fiber) and only make=
a test=0A>> about 60km( mux side A and demux on side B) with sfp+10gb for=
80km... (=0A>> only see ddm on my ex3300( about -19db for 60km). Test swi=
tch access with=0A>> ssh and pinging tests...=0A>> What kind os issue do y=
ou have? For distances less than 60km is this=0A>> solution good?=0A>> Tha=
nks!!!=0A>>=0A>> Enviado via iPhone =EF=A3=BF=0A>> Grupo Connectoway=0A>>=0A=
>>> Em 07/02/2015, =C3=A0s 14:55, Kenneth McRae <kenneth.mcrae@me.com> esc=
reveu:=0A>>> Mike,=0A>>> I just replaced a bunch of FiberStore WDM passive=
muxes with OSI Hardware=0A>>> equipment. The FiberStore gear was a huge d=
isappointment (excessive=0A>>> loss, poor technical support, refusal to is=
sue refund without=0A>>> threatening legal action, etc.). I have had good =
results from the OSI=0A>>> equipment so far. I run passive muxes for CWDM =
(8 - 16 channels).=0A>>> On Feb 07, 2015, at 09:51 AM, Manuel Mar=C3=ADn <=
mmg@transtelco.net> wrote:=0A>>> Hi Mike=0A>>> I can recommend a couple of=
vendors that provide cost effective=0A>>> solutions.=0A>>> Ekinops & Pack=
etlight.=0A>>> On Saturday, February 7, 2015, Mike Hammett <nanog@ics-il.n=
et> wrote:=0A>>> I know there are various Asian vendors for low cost (less=
than $500)=0A>>> muxes=0A>>> to throw 16 or however many colors onto a st=
rand. However, they don't=0A>>> work=0A>>> so well when you don't control =
the optics used on both sides (therefore=0A>>> must use standard wavelengt=
hs), obviously only do a handful of channels=0A>>> and=0A>>> have a distan=
ce limitation.=0A>>> What solutions are out there that don't cost an arm a=
nd a leg?=0A>>> -----=0A>>> Mike Hammett=0A>>> Intelligent Computing Solut=
ions=0A>>> http://www.ics-il.com=0A>>> --=0A>>> TRANSTELCO| Manuel Marin |=
VP Engineering | US: *+1 915-217-2232* | MX:=0A>>> *+52=0A>>> 656-257-110=
9*=0A>>> CONFIDENTIALITY NOTICE: This communication is intended only for t=
he use=0A>>> of the individual or entity to which it is addressed and may =
contain=0A>>> information that is privileged, confidential, and exempt fro=
m disclosure=0A>>> under applicable law. If you are not the intended recip=
ient of this=0A>>> information, you are notified that any use, disseminati=
on, distribution,=0A>>> or=0A>>> copying of the communication is strictly =
prohibited.=0A>>> AVISO DE CONFIDENCIALIDAD: Esta comunicaci=C3=B3n es s=C3=
=B3lo para el uso de la=0A>>> persona o entidad a la que se dirige y puede=
contener informaci=C3=B3n=0A>>> privilegiada, confidencial y exenta de di=
vulgaci=C3=B3n bajo la legislaci=C3=B3n=0A>>> aplicable. Si no es el desti=
natario de esta informaci=C3=B3n, se le notifica=0A>>> que=0A>>> cualquier=
uso, difusi=C3=B3n, distribuci=C3=B3n o copia de la comunicaci=C3=B3n est=
=C3=A1=0A>>> estrictamente prohibido.=
