The IP Over NBMA (ION) Archive

Cell Relay Retreat>ION Archive>month:1996-Jun> msg00140



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Problems with NHC chosen NHS address for registration packets

  • From: jhalpern@us.Newbridge.com (Joel Halpern)
  • Date: Mon, 17 Jun 1996 11:30:19 +0500
  • X-Sun-Charset: US-ASCII

David Horton and Hiroshi Suzuki ask some questions about the
registration packets, and about what should happen under various
error cases.

I had earlier suggested a simple approach which I believe would finesse
the problems.
Reading the current draft, I find that I can not actually tell what
model is proposed by the document.

Speaking simply as a participant in this discussion, allow me to suggest
an approach to the registration packet:
1) In the fixed part of a registration packet, the "Destination Protocol
   Address" is always the IP address of the registering entity.
2) If an NHS which serves the LAG containing that IP address receives
   the request, it either becomes the (a) serving NHS, or sends a NAK
   indicating that service is denied because of resource shortage
3) If an NHS which does not serve the containing LAG receives the 
   registration request, it either forwards it according to routing,
   or establishes a VC back to the source for the purpose of sending
   an error "Unable to forward, no routing path."
This would give an unambiguous procedure, which would whenever possible
result in a relationship with an NHS which is serving the LAG.

To get the ATM address to start the process, the client would need to be
configured with one ATM address.  In most cases, this would be an ATM
Anycast address.
We should allow multiple addresses.

I would be interested in other people's reactions to this suggestion.

Yours,
Joel M. Halpern				jhalpern@newbridge.com
Newbridge Networks Inc.

PS My apologies to the current Author.  The document probably does spell
    out the intended procedures for this.  However, since I want to make
    sure other people get a chance to respond to this, I am sending this
    note out with less the the rigorous re-reading I should apply.