Re: X.25 'user' calling address objects

"Paul S. Rarey" <Paul.S.Rarey@ssf-sys.dhl.com> Tue, 28 January 1992 22:33 UTC

Received: from nri.reston.va.us by NRI.NRI.Reston.VA.US id aa24074; 28 Jan 92 17:33 EST
Received: from dg-rtp.rtp.dg.com by NRI.NRI.Reston.VA.US id aa24015; 28 Jan 92 17:32 EST
Received: from Gateway.DHL.COM by dg-rtp.dg.com (5.4/dg-rtp-proto) id AA16179; Tue, 28 Jan 1992 16:57:10 -0500
Received: by gateway.DHL.COM id ag05497; 28 Jan 92 13:53 PST
Received: from isodev.ssf-sys.DHL.COM by gateway.DHL.COM id aa05383; 28 Jan 92 13:40 PST
To: x25mib@dg-rtp.dg.com
Subject: Re: X.25 'user' calling address objects
In-Reply-To: <9201281703.AA09518@opal.acc.com>
References: <9201281703.AA09518@opal.acc.com>
Organization: DHL Systems Inc., Global Telecommunications Group
X-Mailer: Poste 2.0B2
From: "Paul S. Rarey" <Paul.S.Rarey@ssf-sys.dhl.com>
Sender: prarey@ssf-sys.dhl.com
Date: Tue, 28 Jan 1992 13:40:23 -0800
Message-Id: <920128134023.544@isodev.ssf-sys.DHL.COM>
Encoding: 52 TEXT, 8 TEXT SIGNATURE


> >If the X.25 'MIB Standard' (i.e. the document) does not discuss how users
> >of X.25 should represent their access to X.25 (such as called address) then
> >we promote an inefficient and potentially chaotic and inconsistent collection
> >of variations on this.  This means, in general, that if I at layer N+1 wish
> >to use a service at layer N and define a MIB for myself, I have to invent
> >a way to represent what I send to layer N for addressing information.  This
> >applies to ALL users of X.25 in this specific case.
> 
> As far as internal interfaces, this is obviously an implementation issue.
> As far as external representation, the X25mib defines the x25address
> object type:
> 
>           X25address ::= OCTET STRING (SIZE(0..17))
>            -- 0 to 17 bytes in length containing the ASCII
>            -- characters [0-9], each octet contains one digit
>            -- of the address.

It does not appear that this definition can handle a 15 digit
(30 byte) X.25 (aka X.121) address. Additionally, Called Address
Extensions (Recommendation X.25/1988 Annex G) may not be effectivly
represented with this definition.  

> >Note I am not saying the 'X.25 MIB' has to contain objects for this.  I am
> >suggesting that the document provide advisory implemention for the MIB
> >developers at the layer above.
> > 
> >If the X.25 MIB folks don't define it, who should?  Every single X.25 user-
> >MIB-developer in existance?
> 
> Maybe an example of what you want would be helpful.


Mabey the first byte of the Call User Data (the protocol
identifier) field should be used to denote the "upper layer
protocol" (aka the agent) much the same as RFC-877 uses 0xCC to
denote Internet IP.   This would be preferable from my perspective
since it follows existing conventions and does not consume network
addressing resources. Additionally I would not  relish assigning a
unique network address (X.121, et-al), for application entities.

I agree with the first of Dean's statements...:

> The X.121 address to call for any session belongs to the upper layer
> service asking X.25 to place a call.  To change that address, the
> NMS must use the MIB for the upper layer service.

For the second statment however, I would like to have the option 
to leave address resolution to the servcies that are specificially
implemented for such. 
 



Best regards.....

/psr            Paul S. Rarey
                Open Systems Research and Development
                Global Telecommunications
                DHL Systems Inc.