[c-nsp] LX vs. LH GBICs

[Max Pierson]

Hi Matt,

Here is some output from my Cat6k (IOS) and my Cat4k (CatOS).

Cat4k:
------
1/1  Uplink1	connected	trunk	normal   full  1000 1000-LX/LH
1/2  Downlink1	connected	65    normal   full  1000 1000-LX/LH

Cat6k:
------
GigabitEthernet1/2 is up, line protocol is up (connected)
 Full-duplex, 1000Mb/s, media type is LX

GigabitEthernet3/8 is up, line protocol is up (connected)
  Full-duplex, 1000Mb/s, media type is LH

Both Cat6k links work fine...as well as the Cat4k links.

I would verify the fiber type used from end to end (SMF or MMF). Make
sure your patch cables are the same on your side....this has been known
to cause problems (although I've gotten it to work on special
occasions). Also, you might want to try to disable autonegotiation on
your side to see if the link comes up. This has bitten me a few times
where I could not figure out why the link would not come up...even
though my db levels were in spec.

--nmp


-----Original Message-----
From: Matt Stevens [mailto:matt at elevate.org] 
Sent: Friday, March 21, 2008 7:01 PM
To: Richard A Steenbergen
Cc: Max Pierson; cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] LX vs. LH GBICs

I understand the whole LX/LH concept. I was more wondering what 
specifically the LX/LH GBICs report as their media type (LH?), and if 
there's an older part (or non-Cisco) part that reports LX.

Our connection to a cat65k reporting LH is working, whereas a connection

reporting LX is not.

We've checked levels on this particular link, and everything looks 
within spec for both pieces of equipment - so I'm slightly grasping at 
straws. It's a remote site involving multiple vendors, so 
troubleshooting is painful to say the least.
-- 
matt


Richard A Steenbergen wrote:
> On Fri, Mar 21, 2008 at 03:57:59PM -0500, Max Pierson wrote:
>> 1000BaseLX/LH interfaces are fully comply with the IEEE 802.3z
>> 1000BaseLX standard. However, their higher optical quality allows
them
>> to reach 10 km over single-mode fiber (SMF) versus the 5 km specified
in
>> the standard. This is where the LH kicks in...which allows them to
>> achieve a longer distance when used with SMF.
> 
> But at this point every LX you're ever going to run across does 10km
or 
> better (often much better), and the use of the name "LH" is just a 
> Ciscoism that only serves to confuse people. Also note that different 
> vendors use the names differently, for example Juniper LH is a 70km
1550nm 
> optic (what cisco calls ZX), which is different from its LX 10km
1310nm 
> optic.
> 
> At any rate they're all compatible with each other, all RX units are 
> wide-band, so as long as you aren't trying to engineer something 
> complicated (with a filter, with concerns about dispersion, etc) your
only 
> real concern is "do I have enough optical budget" and maybe "do I need
to 
> attenuate".
>