Re: [NGO] Diff-Serv datamodel for discussion startpoint

[Ladislav Lhotka <lhotka at cesnet.cz>]

OKITA Hideki píše v Út 20. 03. 2007 v 22:31 +0900:
> At the OPS are meeting in Wednesday, Iijima-san will talk about 
> operators/developers' viewpoint to the NETCONF architecture along
> this subject.

Yup, I'll be there.

> 
> 
> 
> > I fixed one minor thing that IMO is not correct in the XSD schema  - if
> > the SrcPrefixLength and DstPrefixLength elements (in classifierUnit) are
> > specified, then the respective SrcAddress nad DstAddress must also be
> > present.
> 
> Does it mean that a Src/DstPrefixLength element should be subordinate
> to a Src/DstAddress element in a <classifierUnit> element to avoid
> independent appearance of a Src/DstPrefixLength element?  If so,
> I can agree with you.

Yes, each address/prefix pair is optional and within it the prefix is
also optional. 

> 
> However, it is good to change the expression about IP address
> to follow the Andy's ipAddress module. The present expression follows
> the "DiffServMultiFieldClfrEntry" object defined in the Diff-Serv MIB.
> For such basic data types, IMHO, we do not need to follow SMI-style.

There are few other data modelling details that could be worth
discussing but for the sake of comparing the schema languages these
examples are quite characteristic.

Lada

-- 
Ladislav Lhotka, CESNET
PGP Key ID: E74E8C0C


_______________________________________________
NGO mailing list
NGO at ietf.org
https://www1.ietf.org/mailman/listinfo/ngo