From: MADMAN (dmadlan@xxxxxxxxx)
Date: Wed Mar 20 2002 - 14:02:18 GMT-3
A big differance is a UNI only responds to LMI requests, never
initiates them whereas an NNI initiates and responds. This is crucial
when a customer has a PVC that traverses providers, if a PVC goes down
in provider A provider B will learn this and pass it along so that your
PVC will go inactive. An NNI also is how Lucent talks to Stratacom,
etc... but within our own frame network, all Lucent those connections
are called trunks and run proprietary protocols.
ATM NNI's allow 1024 VP and 65K VC, UNI's allow 64 VP and 65K VC'c
Dave
Jason Wydra wrote:
>
> Could someone tell me the difference between a UNI and NNI in Frame Relay. I
understand that UNI is used between the CPE and the provider switch and that NN
I is used between two different provider switches. My question is, what is actu
ally different in the frame packet or what signaling techniques are used to dif
fer the two? I understand that adding ANSI annex D allows both devices at eithe
r end to both initiate and respond to status messages/keepalives simultaneously
. So why not just run UNI with Annex D everywhere and forget about having NNI a
ltogether. Do we need to put a label on frame connections so that we know which
one is a link between a CPE and switch and which one is a link between two swi
tches? Am I way off base????
>
> Furthermore - The only difference I see between an ATM UNI and NNI is the GFC
field and and extended VPI field. Big deal? Why have two separate specificatio
ns?
>
> Thanks,
>
> Jason (NP,DP)
>
> ---------------------------------
> Do You Yahoo!?
> Yahoo! Sports - live college hoops coverage
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:57:14 GMT-3