Hi!
I seem to suffer from an acute lack of 10GE CWDM optics options. Is it just me or am I just looking in all the wrong places?
You'd think that by now there would be an upgrade market from 1GE to 10GE. DWDM wavelenghts are not always available, but CWDM often are.
- Zed
These are XFP-based. Thus, not a solution to the problem above.
Answer: Nobody's making 10GE CWDM-wavelength lasers. Why? I don't have
enough knowledge of optical equipment, but my understanding is that
it is because:
a) Currently DWDM component suppliers already have a full load of
orders and have problems scaling production - as evidenced by long
lead times on any DWDM optics.
b) They wouldn't be much cheaper to produce than
temperature-stabilized DWDM optics.
c) The demand is currently for amplifiable DWDM optics.
-alex
Zed,
If you are looking for optical systems my fav pub is Lightwave at http://lw.pennnet.com/. They list DWDM and CWDM systems, lasers, optics, ROADM etc. If you have nanog archives go back at least six months to the thread on DWDM vs CWDM et al that I was one of the contributors to.
The last comment you make I do not understand since:
1. DWDM systems have many more lasers / wavelengths that are used and most metro and WAN providers can supply you a ITU wavelength at the normally obscene prices.
2. CWDM systems are usually 4, 8, 16, 24 that are also on ITU wavelengths like DWDM systems but with 50 - 100 nm or more spacing so the lasers and optics do not have to be so precise as the DWDM optics.
Currently it is my understanding the 10 Gbps signals are carried on 4 x 2.5 Gbps signals that are compatible with existing CWDM and DWDM equipment. There are 40 Gbps DWDM systems and 10 Gbps lasers on 100 Gbps and greater capacity systems. I agree with Alex's comments that to have 10 Gbps on a CWDM system is to have a CWDM system of at least 40 to 100 Gbps and that is very expensive today.
John (ISDN) Lee
From Lightwafve:
Optelian adds CWDM XFP transceivers
AUGUST 19, 2008 -- Optelian (search for (search for Optelian)) has announced the availability of new LightGAIN 10-Gbit/sec CWDM XFP transceivers, which the company claims create a cost breakthrough in 10-Gbit/sec capacity growth by combining the low price points of CWDM with the inventory cost benefits of using MSA-compliant pluggable transceivers.
"These new CWDM XFP transceivers are a cost-saving solution for fiber-constrained customers looking to grow their 10-Gbit/sec services [but] who don't need the full migration path provided by DWDM," explains Dave Dal Farra, senior Optelian product manager. "And in higher growth networks, the low first cost and inventory savings can still be taken advantage of by adding up to 9 DWDM wavelengths into unused CWDM channels, utilizing Optelian's hybrid CWDM/DWDM multiplexers and 10-Gbit/sec tunable DWDM regenerators."
The new CWDM XFP transceivers are fully supported and backwards compatible in LightGAIN 6140, 5140, and 3060 systems, plugging into the existing RGN-10GXF 10-Gbit/sec regenerator card. As part of LightGAIN, the flexibility of CWDM pluggable transceivers combines with Optelian's Quick-Turn Custom Passives so that custom 10-Gbit/sec CWDM configurations can be delivered as quickly as two weeks, claim company representatives.
Also useful for wavelength conversion, reach extension, and regeneration, the 10-Gbit/sec CWDM XFP transceivers are now available for order. Key specifications include coverage from 1471 nm to 1611 nm, plus SONET and Gigabit Ethernet compliance for data rates with or without G.709 FEC.
From Alex's older e-mail:
Subscribe to Lightwave (at no charge) and look at the back issues for networks. Show up at Supercom or OFC or what is replacing them and get the latest on ROADM, full channel tunable lasers and maintenance costs.
What size of network do you want to grow to before replacing the optical link equipment including ILAs?
Most any org can cost justify a CWDM / CAN since you can add one fiber pair at a time and one lambda per fiber pair.
DWDM gear is much more expensive and is aimed at 20 to 40 lambdas per
fiber for service providers while UDWDM and ULHWAN are aimed at trans
oceanic links and are very very expensive.
DWDM gear is not expensive. Passive muxes cost little. Active
transceivers cost money but not very expensive at all.
Check out these two presentations (by yours truly et al):
http://www.nanog.org/mtg-0606/pdf/lightning-talks/4-pilosov.pdf
http://www.nanog.org/mtg-0610/presenter-pdfs/pilosov.pdf
-alex
The only affordable CWDM 10G system I have seen although I haven't used it yet is a single 10G band at 1310 or 1550 with 8 additional 2.5G bands around it. I haven't seen any 4 band 10G CWDM boxes with XFPs for less than $5000 yet, but I would expect them in the next year or two - I'm hoping anyway. I'm out of the country at the moment and access is a bit too slow to look it up easily now. If you need the manufacturer, let me know and I'll look it up when I return.
-Robert
Tellurian Networks - Global Hosting Solutions Since 1995
http://www.tellurian.com | 888-TELLURIAN | 973-300-9211
"Well done is better than well said." - Benjamin Franklin
Depending how cheap and ghetto you want to get, there's also possibility
of doing WDM on 1310/1300. I have custom-manufactured splitters filtering
1307nm +-2nm - and any given LR XFP [*1] will be either within that band
or outside [*2]. Test a bunch of them, split them into two groups, use on
the "tested" wavelength. Bunch of friends&family are using this technology
in production. This gives you an ability to do 20G with very cheap optics.
[*1] Except ones with very temperature dependent wavelength - mark them as
"warms up to 1300" and use if you don't care that your links will take
about 5 minutes to "warm up" and come up. 
[*2] Any LX4 Xenpak would be "outside" of the band as well, and you can
use LX4 concurrently with LR.
There are some more ghetto fabulous things you can do, described in
http://www.nanog.org/mtg-0610/presenter-pdfs/pilosov.pdf 
-alex
Hello Alex:
Depending how cheap and ghetto you want to get, there's also possibility
of doing WDM on 1310/1300. I have custom-manufactured splitters filtering
1307nm +-2nm - and any given LR XFP [*1] will be either within that band
or outside [*2]. Test a bunch of them, split them into two groups, use on
the "tested" wavelength. Bunch of friends&family are using this technology
in production. This gives you an ability to do 20G with very cheap optics.[*1] Except ones with very temperature dependent wavelength - mark them as
"warms up to 1300" and use if you don't care that your links will take
about 5 minutes to "warm up" and come up.[*2] Any LX4 Xenpak would be "outside" of the band as well, and you can
use LX4 concurrently with LR.There are some more ghetto fabulous things you can do, described in
http://www.nanog.org/mtg-0610/presenter-pdfs/pilosov.pdf-alex
Do you have any issues with four wave mixing or other crosstalk issues or do you account for this in your channel plan?
Regards,
Mike
The only affordable CWDM 10G system I have seen although I haven't
used it yet
> is a single 10G band at 1310 or 1550 with 8 additional 2.5G bands
around it.
I've wondered if one could shoot with DWDM 10G optics into two channels
of a CWDM mux. For example, by connecting DWDM channel 359 (center
1530.33 nm) and 334 (center 1550.12 nm) to the 1530/1550 filters of a
CWDM mux with 20nm spacing (+/- 6.5nm pass band). Might that support
1x10gig + 3x1gig on a single strand, or 2x10G + 6x1G on a pair? (and
no, I haven't tried it).
Bradley
CWDM filter bandpass is wide to allow for drifting optics. Anything
within about 7nm of 1530/1550 should work fine. I've got some optics
near 34 and 59 on order to do exactly that in a bidir single fiber
arrangement. I'll report back my results.
Yes, that'll work fine.
You can also use DWDM mux to mux down 1529.53/1530.33/1531.13 and put them
all into 1530 channel. (Same with 1550 channels).
http://www.nanog.org/meetings/nanog38/presentations/pilosov.pdf
http://www.nanog.org/meetings/nanog37/presentations/4-pilosov.pdf
