RE: Plug-and-play components (was FW: Single vs many solution(s))

Ok, so now I get it.

Its all about the architectural ability to use different components -I
didn't get that originally.

More offline.

Mark

-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Tuesday, May 06, 2003 3:58 PM
To: [email protected]
Cc: [email protected]
Subject: RE: Plug-and-play components (was FW: Single vs many
solution(s))


Hi Mark,
> Neil,
>
> [clipped]
>
> >> As an operator I have a requirement to be able to
> choose addressing, routing, signalling (for the 2 co
> modes), OAM, etc functional components *independently*
> (using best-of-breed selection criteria) both per
> mode and within each mode.<<
>
> Not sure how this can be achieved in practice (under
> current industry structure/system architectures).
NH=> Easy....just find a vendor who recognises it!  We did for L1
control-plane signalling and are now happily using PNNI (as we wanted).  It
a great signalling protocol...better than rsvp-te in our view.  I know
another large operator who has done exactly the same as we have (even bought
from the same vendor) though is still trotting out the GMPLS stuff on the
lists....quite bizarre.  As I noted in an earlier mail....its not at all
clear to me that those vocal on the IETF lists represent an operator's true
internal position.

>