[j-nsp] 3rd party DWDM SFPs

Richard A Steenbergen ras at e-gerbil.net
Thu Jun 12 14:31:03 EDT 2008


On Thu, Jun 12, 2008 at 10:53:31AM +0200, sthaug at nethelp.no wrote:
> You can get a *little* more info from "show chassis pic":

Ah didn't know about that one.

> PIC port information:
>                           Fiber                    Xcvr vendor
>   Port  Cable type        type  Xcvr vendor        part number     Wavelength
>   0     GIGE 1000LH       SM    FIBERXON INC.      6128C-SL5051G   1510 nm
>   1     GIGE 1000LX       SM    CISCO-FINISAR      1000BASE-LX     n/a
>   2     GIGE 1000LX       SM    CISCO-FINISAR      1000BASE-LX     n/a
>   3     unknown cable     SM    CISCO-FINISAR      F-WDM-1530      n/a

Strangely enough (this is on MX960), it reads the wavelength value correct 
in the PFE shell but does't display it on the CLI for me:

PIC port information:
                          Fiber                    Xcvr vendor
  Port  Cable type        type  Xcvr vendor        part number       Wavelength
  0     unknown cable     n/a   FINISAR CORP.      FTRX-3811-328-TR  n/a     

For some LRs:

  0     10GBASE LR        n/a   FINISAR CORP.      FTRX-1411-3       n/a     
  0     10GBASE LR        n/a   CISCO-OPNEXT       TRF5011AN-LF004   n/a     
  0     10GBASE LR        n/a   CISCO-AVAGO        QFCT-7081         n/a     

On Thu, Jun 12, 2008 at 10:41:25AM -0400, Chuck Anderson wrote:
> Can you get digital diagnostics info from these too?  Here's the info
> from one of our standard Juniper SFPs.  Note that "show interfaces   
> diagnostics optics" only works if you give it an interface argument  
> (including wildcards).  Otherwise it shows the standard "show
> interfaces" information for all ports, but no digital diagnostics
> info.

Yeah I've been complaining about that one since at least Jan 07, I don't 
think anyone cares to fix it. :) I've never found any XFP optics which 
don't correctly do DOM on a Juniper at any rate.

One weird thing to watch out for, Juniper doesn't correctly calculate dBm 
for the optical power levels in the case where the power is 0 mW.

Physical interface: xe-1/0/0
    Laser bias current                        :  31.176 mA
    Laser output power                        :  0.5720 mW / -2.43 dBm
    Module temperature                        :  35 degrees C / 95 degrees F
    Laser rx power                            :  0.5205 mW / -2.84 dBm

What used to happen when you had no detectable signal coming in was you 
would get:

    Laser rx power                            :  0.0000 mW / 0.00 dBm

Clearly this is incorrect, since 0 dBm is 1 mW (a strong signal). The 
correct value for dBm in the case of 0 mW is negative infinity. In new 
code they've "corrected" this by just not displaying dBm (or any units at 
all) when the RX power is 0:

    Laser rx power                            :  0.0000

They also skip the XML field completely when this happens, so you have to 
parse around it:

ras at router> ...s optics xe-1/0/0 | display xml | match rx-optic
                <laser-rx-optical-power>0.5205</laser-rx-optical-power>
                <laser-rx-optical-power-dbm>-2.84</laser-rx-optical-power-dbm>

ras at router> ...tics optics xe-0/3/0 | display xml | match rx-optic
                <laser-rx-optical-power>0.0000</laser-rx-optical-power>

But I've actually seen cases where people have been looking at the dBm 
field only and said "0dB thats a strong signal, this must be fine". :)

-- 
Richard A Steenbergen <ras at e-gerbil.net>       http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)


More information about the juniper-nsp mailing list