[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Isis-wg] Error in IANA TLVs listing



I'm not really sure how that would help unless we could get the names changed in ISO/IEC 10589 itself. I rather doubt both the ability to do that, and the desirability, even if we could.

We would simply be replacing the potential confusion caused by having two TLVs with the same name, but different numbers and which are used in different PDUs, with the confusion caused by having a TLV (2) which is called "Intermediate System Neighbors" in ISO/IEC 10589 (and probably much documentation and code based on that) and called "IS reachability TLV" in some obscure RFC.

Incidentally this duplicate naming has persisted since the days of DECnet PhaseV, where they were both called "Router Neighbors", since that is what they are. They contain the information about router (or in ISO terms IS) neighbors as is appropriate to the PDU in which they appear. Since they had different type numbers and appeared in different PDUs, there didn't seem any confusion.

I agree, however, that RFC 3563 is in error and the "name" field therein should be changed to match the relevant standard(s).

I haven't done so, but it would be wise to check the remaining names for consistency as well.

   Mike

Donald Eastlake wrote:
I'd be happy to write an RFC to do this if it seems like it has a
reasonable probability of approval. I'll poke around a bit and see if
there seems to be anything else such a hypothetical "IS-IS TLV
Nomenclature" RFC should do.

Thanks,
Donald

On Wed, Sep 9, 2009 at 10:42 PM, Tony Li<tony.li at tony.li> wrote:
Donald,

If you want to propose name changes and fix the overloading, that would be
fine by me.  However, I believe that would properly require an RFC.

Tony


Donald Eastlake wrote:
OK, I'll send in an errata to IANA.

I think maintaining conflicting names is a bad idea so, while of
course the expert can do whatever they like, I plan to recommend that
the main listing name for TLV #2 be "IS Reachability TLV" with a
footnote stating that the ISO standard confusingly calls it the same
thing as TLV #6.

Thanks,
Donald

On Tue, Sep 8, 2009 at 5:59 AM, mike shand<mshand at cisco.com> wrote:
I agree with Tony.

 Mike

Tony Li wrote:
Hi Donald,

Donald Eastlake wrote:
Hi,

I believe there is a significant error in the IANA listing of IS-IS
TLVs
at
http://www.iana.org/assignments/isis-tlv-codepoints
I confirm the error.  It seems to source from RFC 3359, which initially
populated the registry.


In particular, both TLV #2 and TLV #6 are listed as "IIS Neighbors". I
believe that is correct for TLV #6 but that TLV #2 is the older IS
Reachability TLV ("narrow metric"). Should someone contact IANA to get
this fixed?

PS: Actually, to be persnickety, since the reference is ISP 10589,
looking there, I think TLV #6 is called "IS Neighbors" or
"Intermediate System Neighbors", not "IIS Neighbors".
TLV 6 is indeed "Intermediate System Neighbors" (see 9.6).  However, TLV
2
is also "Intermediate System Neighbors"  (see 9.8, 9.9).

I believe that the correct procedure here is for someone, in this case
you, as discoverer, to contact IANA and inform them of the error.  As
the
registry is managed under the 'expert review' process, they can contact
their experts and have this confirmed and corrected.

I propose that the correction be to make them both "IS Neighbors".  It's
unfortunate that 10589 overloaded the term, but I believe that we're not
in
a position to correct that at this point.
Regards,
Tony




_______________________________________________
Isis-wg mailing list
Isis-wg at ietf.org
https://www.ietf.org/mailman/listinfo/isis-wg

_______________________________________________
Isis-wg mailing list
Isis-wg at ietf.org
https://www.ietf.org/mailman/listinfo/isis-wg