I apologize for the delayed response: I'm trying to catch up with my
email.

The thing that bothers me about this otherwise excellent discussion is
that I interpret these emails as implying that all identifiers must be
known at the IP layer. A session ID is a transport layer or above
concept. It is desirable (e.g., for management and security) to be able
to correlate session IDs with locators and also with physical locations,
but this doesn't mean that it is desirable to map these identifiers
together within the same layer. Put another way, it is fine for locators
to be in the IP layer and identified at that layer and session IDs to be
at a transport and above layer and identified there. Should a function
need to correlate the two then their existence within their appropriate
layers would permit that in the general case.

-----Original Message-----
From: Scott Brim [mailto:[email protected]] 
Sent: Wednesday, July 15, 2009 11:06 AM
To: Lixia Zhang
Cc: [email protected]
Subject: Re: [rrg] Next topic: properties of identifiers

Lixia Zhang allegedly wrote on 07/15/2009 1:52 PM:
>> Yes, we absolutely need this.  But why does this imply the need for a

>> host or stack identifier?  Session identifiers already map to the set

>> of locators that go to the same communicating entity.  This is why we

>> need to distinguish them from service identifiers, which in turn may 
>> map to locators of different communicating entities.
> 
> the above comment seems brining more implicit assumptions regarding 
> "session identifiers", e.g. "Session identifiers already map to the 
> set of locators that go to the same communicating entity..."

This mapping of session identifier -> locators is transient and not for
general use.  That is, you couldn't put a session identifier in DNS and
expect to use it as a "where is this node" locator.  However, while a
session exists, participants in that session can use a session
identifier internally.

> as I mentioned in the prev msg, I've yet to see a simple and clear 
> definition of "Session" first.  As Joel questioned in next msg,

See previous reply.

>    I am trying to refer someone to a specific entity. They
>    do not have a session with that entity.  So a session ID
>    is clearly totally useless for a referral.

That's a different purpose and the statement is correct.  A session ID
is meaningless if there is no session.
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg

Reply via email to