Re: [Teas] Last Call: <draft-ietf-teas-yang-te-types-09.txt> (Traffic Engineering Common YANG Types) to Proposed Standard

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

 



<at the end>

Tom Petch

----- Original Message -----
From: "Tarek Saad" <tsaad.net@xxxxxxxxx>
Sent: Wednesday, May 15, 2019 10:13 PM

> Hi Tom,
>
> Thanks for sharing your feedback.
> I'm attaching Igor's response on this topic -- which I share his same
opinion.
> Please see more comments inline from me [TS]..
>
> On 5/15/19, 7:16 AM, "Teas on behalf of tom petch" wrote:
>
>     The approach taken by this I-D worries me.
>
>     It provides YANG identities for a wide range of values used in TE,
such
>     as encoding types and switching capabilities; so far, so good.
> [TS]: As mentioned in abstract, the module is not strictly identities.
It is a collection of re-usable YANG groupings, types and identities.
>
>     These definitions were needed, and were in a large part created by
>     RFC3471, in 2003.  When the management of GMPLS was specified, in
MIB
>     modules, these definitions were put under IANA control and they
remain
>     there to this day.  They were updated by e.g. RFC8330 (February
2018)
>     and RFC8363 (May 2018) so these IANA registries are not some dusty
old
>     relic but a current, living specification.
>
>     These YANG definitions have much in common with the IANA SMI
registries
>     but they are not the same.  A comparison of e.g. switching
capabilities
>     suggests that this YANG module is out-of-date compared with the
IANA SMI
>     registry (as with RFC8330, RFC8363) and omits several values for
no
>     stated reason ( the deprecated 2,3,4, 40 PBB-TE, 151 WSON-LSC).
> [TS]: it was not the intention to be exhaustive in covering all IANA
defined entities. However, the objective was to model enough that would
make TE feature (modelled in other modules) usable and to leverage the
power of YANG augmentation for any extensions that may not be covered in
a base model. Specifically, the authors favored the use of YANG
identities over enums to allow for the extensibility of augmentation.
>
>     The approach taken by other WG has been to take a IANA registry
and
>     provide a parallel YANG module under common IANA control as has
been
>     done for e.g. interfaces with both MIB module and YANG module
being
>     updated in parallel as appropriate.
>
>     Here something seems to have gone wrong.  We have a parallel set
of
>     definitions not acknowledging the existing ones and being
out-of-date
>     compared with the existing ones.
>
>     Furthermore, some of these definitions are duplicated in the work
of the
>     LSR WG giving us (at least) three definitions.
> [TS]: it would help to point to the duplication or thread that this
was discussed in. However, we believe that this document covers TE data
and hence LSR module(s) would need to eliminate duplication of any TE
data (if any).. LSR module can always import TE types to use the TE
definitions.
>
>     I raised this issue before Christmas 2018 and was told that the
chairs
>     of TEAS and LSR would get together and get back to me.  Nothing
appears
>     to have changed.
>
>     In passing, IANA has separate SMI registries for e.g.LSP encoding,
>     Switching Types and so on, which seems a sound engineering
approach,
>     allowing more flexible evolution compared to the 60-page monolith
of
>     this single YANG module.
> [TS]: In this effort, we've followed similar approach to RFC8294,
RFC6021. Etc.. Do you see the same concerns there too?

I think that the difference is the sheer size of this I-D and the likely
future evolution of it.  The IANA SMI registries each form a logical set
each of which can proceed independently and I think that that is spot
on; RFC8294 is much smaller and so less of a concern, and is also more a
loose aggregation of related objects  with no natural subdivisions.

Here you have replicated the objects in.
  identity lsp-encoding-types {
exactly AFAICT and so that is an obvious candidate to turn the SMI into
SMI and YANG.

With
  identity switching-capabilities {
the SMI is more up-to-date than the YANG and I would see  the SMI as the
starting point

With
 identity wdm-spectrum-type
you have replicated the objects but seem to have got the reference
wrong; you have
  identity flexible-grid { base wdm-spectrum-type; description "Flexible
grid.";
    reference "RFC6205";
whereas SMI has
3, ITU-T Flex, [RFC7699]
which I think correct (now if you had used the SMI as a basis, .....:-)

Tom Petch

>
> Regards,
> Tarek
>
>     ..Tom Petch
>
>     ----- Original Message -----
>     From: "The IESG" <iesg-secretary@xxxxxxxx>
>     Sent: Thursday, May 02, 2019 9:47 PM
>
>     > The IESG has received a request from the Traffic Engineering
>     Architecture and
>     > Signaling WG (teas) to consider the following document: -
'Traffic
>     > Engineering Common YANG Types'
>     >   <draft-ietf-teas-yang-te-types-09.txt> as Proposed Standard
>     >
>     > The IESG plans to make a decision in the next few weeks, and
solicits
>     final
>     > comments on this action. Please send substantive comments to the
>     > ietf@xxxxxxxx mailing lists by 2019-05-16. Exceptionally,
comments may
>     be
>     > sent to iesg@xxxxxxxx instead. In either case, please retain the
>     beginning of
>     > the Subject line to allow automated sorting.
>     >
>     > Abstract
>     >
>     >
>     >    This document defines a collection of common data types and
>     groupings
>     >    in YANG data modeling language.  These derived common types
and
>     >    groupings are intended to be imported by modules that model
Traffic
>     >    Engineering (TE) configuration and state capabilities.
>     >
>     >
>     >
>     >
>     > The file can be obtained via
>     > https://datatracker.ietf.org/doc/draft-ietf-teas-yang-te-types/
>     >
>     > IESG discussion can be tracked via
>     >
https://datatracker.ietf.org/doc/draft-ietf-teas-yang-te-types/ballot/
>     >
>     >
>     > No IPR declarations have been submitted directly on this I-D.
>





[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Mhonarc]     [Fedora Users]

  Powered by Linux