Re: [MIB-DOCTORS] [IPFIX] AD review of draft-ietf-ipfix-mib-05.txt(ipfixExportEntry)

["Randy Presuhn" <randy_presuhn at mindspring.com>]

Hi -

> From: "Atsushi Kobayashi" <akoba at nttv6.net>
> To: "Thomas Dietz" <Thomas.Dietz at nw.neclab.eu>
> Cc: "IETF IPFIX Working Group" <ipfix at ietf.org>; "MIB Doctors (E-mail)" <mib-doctors at ietf.org>
> Sent: Thursday, May 07, 2009 10:46 PM
> Subject: Re: [MIB-DOCTORS] [IPFIX] AD review of draft-ietf-ipfix-mib-05.txt(ipfixExportEntry)
...
> > > T16. ipfixExportEntry leads to a warning in the smicng compilation
> > > 
> > > W: f(ipfix.mi2), (676,12) Row "ipfixExportEntry" does not have a
> > > consistent indexing scheme - index items from current table must come
> > > after index items from other tables
> > > 
> > 
> > Since I have only libsmi (which does not catch this) I did not realize it.
> > Fixed.

This smicng warning can be misleading.  There is no SMI requirement, nor is
there even a recommendation, for having "a consistent indexing scheme."
The order of indexes should be determined by:
  (1) the use cases for retrieving subsets of the table
  (2) ease of access control policy formulation.

Making index order "consistent" can potentially defeat both of these
objectives.

Randy