RE: [T1X1.5] RE: LCAS and draft-mannie-ccamp-gmpls-lbm-tdm-00.txt

["Grant W. Cyboron" <cyboron@paramanet.com>]

Mannie,

The difficulty that I see in using the C2 byte for synchronizing the
switchover to the increased bandwidth lies in the fact that the source
will have no knowledge of when the sink has accepted the new value of C2
as valid. G.806 section 6.2.4.2 specifies that the sink accepts the new
value of C2 in m frames where 3=<m=<10. Thus there could be up to a 7
frame difference in when the source expects the sink to have accepted
the C2 value and when it actually has (a total of 875 microseconds). I
would expect this to be considered a "disruptive event."  

Grant W. Cyboron
Parama Networks
cyboron@paramanet.com <mailto:cyboron@paramanet.com> 
+1 732 544 9120, x710 (voice)
+1 732 544 9121 (fax)


-----Original Message-----
From: Mannie, Eric [mailto:Eric.Mannie@ebone.com]
Sent: Monday, November 19, 2001 12:12 PM
To: 'mvissers@lucent.com'
Cc: 'Stephen Trowbridge'; ccamp; q11/15; t1x1.5; Dimitri Papadimitriou
Subject: [T1X1.5] RE: LCAS and draft-mannie-ccamp-gmpls-lbm-tdm-00.txt


Hello Maarten,

More about your e-mail:

>Looking at your simple example, I get the impression that you do not
fully
>understand the use of the UNEQ value in C2/V5. The UNEQ value in C2/V5
is
>generated by the HOVC/LOVC connection function (see G.783) as part of
the
UNEQ
>VC-n signal generation process.

Good, that's exactly what we want and need with GMPLS-LBM !

>The UNEQ value is not generated by the path
>termination function or its associated adaptation function. An
adaptation
>function alsways generates an equipped-specific signal label (>= 2).

Good, that's exactly what we want and need with GMPLS-LBM !

We never said on which specific functions acts GMPLS-LBM. At one point
in
the time the adaptation function has to generate an equipped signal, we
just
said that we need to control that event.

The triggering of C2 is done in ONE frame, in LCAS you MUST wait for
multiple frames. So, what do you conclude for the LBM performance?

By the way, this is draft 0 and we could even think at using other C2
transitions as a trigger, like first using a test signal and then
switching
to the real data signal. The issue is to find the most appropriate
transition or even the list of the possible transitions.

>I.e. the assumptions underlying your steps 2 to 6 are incorrect and not
aligned
>with the SDH specifications in ITU-T.

It depends what YOU understood from GMPLS, on-going GMPLS protection and
restoration work, and GMPLS-LBM of course.

Kind regards,

Eric