1000BaseSX to 100BaseFX

Archived from groups: comp.dcom.lans.ethernet (More info?)

I recently checked out one of our networks to find that our switches
with 1000BaseSX uplinks were connected to, and functioning with, our
100baseFX switches. Also, from what I've been told and seen (from the
brief time that I've been here) it has been working this way without
errors, for some time now.

My question is: how is this working if one side is gigabit and the
other is 100 mbit? (please include gory details)

Thanks

Archived from groups: comp.dcom.lans.ethernet (More info?)

"M.C. van den Bovenkamp" <marco@linuxgoeroe.dhs.org> wrote in message
news:41b101ae$0$73696$e4fe514c@dreader19.news.xs4all.nl...
> Tom wrote:
>
> > I recently checked out one of our networks to find that our switches
> > with 1000BaseSX uplinks were connected to, and functioning with, our
> > 100baseFX switches. Also, from what I've been told and seen (from the
> > brief time that I've been here) it has been working this way without
> > errors, for some time now.
> >
> > My question is: how is this working if one side is gigabit and the
> > other is 100 mbit? (please include gory details)
>
> ??? Are you *very* sure about that? 100BaseFX & 1000BaseSX shouldn't
> even be able to see each other, much less talk to each other!
>
> And I do mean 'see' literally: 100BaseFX uses a wavelength of 1300 nm,
> and 1000BaseSX uses 850 nm. They shouldn't even detect each others'
> presence.

I have heard rumours about GBICs with 100 base-FX fibre interfaces - maybe
that is what is in the switch?
>
> On that front you'd have more luck with -LX, which also uses 1300 nm,
> but even then I have never heard of a 1000BaseLX PHY that does 100BaseFX
> as a fallback...
>
> Switch make & model? Module types? Inquiring minds want to know :-).
>
> Regards,
>
> Marco.
--
Regards

Stephen Hope - return address needs fewer xxs