The IP Over NBMA (ION) Archive

[Grenville Armitage <gja@dnrc.bell-labs.com>]

Hi Sebastien,

The CSN is specific to a particular instance of a MARS Cluster.
A MARS can support (in theory at least) multiple Clusters, one
per L3 protocol. So yes, the CSN is interpreted in the context
of the protocol referred to in the message containing the
sequence number.

gja

Sebastien Faure wrote:
> 
> want my MARS to support multiple layer 3 protocol. And something is
> unclear to me about the CSN (Cluster Sequence Number).
>         In Section 9 of RFC 2022, it is said that "Every MARS and MARS
> CLient MUST treat CMI in the context of the protocol type carried in the
> MARS message", but what about the CSN?
> 1 Should one consider it the same way as CMI, that is to say, a MARS MUST
> manage as CSNs as it supports layer 3 protocol, and each MARS Client must
> interpret the CSN in the context of the protocol type carried in the MARS
> message?
> 2 Should one consider a unique CSN for a given MARS? And thus the
> following
> situation can happen:
> A MARS Client will try to revalidate all the VCs of its caches when it
> misses a MARS
> message even if the missed message didn't concern a protocol that this
> client support
> 
> Is it a implementator choice?
> 
> Thanks
> 
> Sebastien FAURE

-- 
________________________________________________________________________
Grenville Armitage                          High Speed Networks Research
http://nj5.injersey.com/~gja              Bell Labs, Lucent Technologies