Re: [Isis-wg] I-D Action:draft-ietf-isis-layer2-05.txt

[Donald Eastlake <[e-mail]>]

Hi Les,

Thanks for this comment. I agree with your arguments and believe that
there is insufficient reason for a separate PDU number for TRILL
Hellos. TRILL Hellos should just use the LAN IIH PDU number.

Donald
=============================
 Donald E. Eastlake 3rd   [phone] (cell)
 155 Beaver Street
 Milford, MA 01757 USA

On Thu, May 6, 2010 at 1:52 AM, Les Ginsberg (ginsberg)
<[e-mail]> wrote:
> Donald -
>
>> -----Original Message-----
>> From: [e-mail] [mailto:[e-mail]] On
>> Behalf Of Donald Eastlake
>> Sent: Wednesday, May 05, 2010 6:29 AM
>> To: Ralph Droms
>> Cc: [e-mail]
>> Subject: Re: [Isis-wg] I-D Action:draft-ietf-isis-layer2-05.txt
>>
>> Hi Ralph,
>>
>> The existing table in draft-ietf-isis-layer2-05 in Section 5, split
>> across pages 56 and 57, has an "LSP" column and an "MGROUP LSP"
> column.
>> By implication, those protocols indicated in the TRILL/IEEE/OTV column
>> for an TLV/sub-TLV with an X in the MGROUP LSP column are specified in
>> that draft as using the MGROUP set of PDUs. However, I, along with
> many
>> others, think the MGROUP cluster of three PDUs is unnecessary and
>> should just be eliminated from the draft.
>>
>> That leaves three other PDUs, the TRILL Hello, MTU-probe, and MTU-ack.
>>
>> Not surprisingly, the "TRILL Hello" PDU is used by TRILL but it is a
>> separate PDU only out of an abundance of caution. TRILL
> implementations
>> never use the "LAN Hello" used at Layer 3 and non-TRILL
> implementations
>> of IS-IS never issue a TRILL Hello. (Both use the same P2P Hello on
> P2P
>> links.) Since TRILL and Layer 3 frames are distinguished by using
>> different multicast addresses and by using Area Addresses which are in
>> practice, if not in theory, completely disjoint. Thus there are
> already
>> 2 means of separating "LAN Hellos" frames from "TRILL Hellos" and, as
> I
>> say, using a different PDU number for TRILL Hellos was suggested only
>> out of an abundance of caution.
>
> If such caution is needed for LAN IIHs, why is it not also needed for
> P2P Hellos, LSPs, CSNPs, and PSNPs?
>
> The answer that because there is new TLV information in TRILL Hellos
> seems inadequate as there is clearly also new TLV information in LSPs
> when used for L2.
>
> I would also point out that in
> http://tools.ietf.org/id/draft-ietf-isis-mi-02.txt wherein multiple
> instances of Layer 3 would be operating on the same link, no new PDUs
> are specified.
>
> The presence of different multicast destination addresses seems quite
> robust. If it is known that both L2 and L3 are operating on the same
> link different system IDs can be used for the two instances. This
> provides additional protection in the case that a neighboring legacy L3
> instance mistakenly processes L2 PDUs. Since the L2 instance will not
> form an adjacency w the L3 instance, no two way connectivity will ever
> be achieved.
>
> I see no need for a new PDU for "TRILL Hellos".
>
>   Les
>
>
>>
>> The MTU-probe and MTU-ack provide a simple one-hop request-response
> way
>> to test link MTU. Implementation of MTU-probe is optional but a TRILL
>> implementation receiving an MTU-prove must respond with an MTU-ack.
> MTU
>> is a loop safety consideration for TRILL. This MTU facility, and the
>> ability to report MTUs with the extended IS reachability MTU sub-TLV,
>> could also be used, for example, by Layer-3 IS-IS to collect
>> information on which to base traffic engineered routes meeting some
>> minimum MTU requirement.
>>
>> Thanks,
>> Donald
>>
>> On Wed, May 5, 2010 at 8:36 AM, Ralph Droms <[e-mail]>
>> wrote:
>>
>>
>>       John - I understand that OTV is a proprietary protocol.  My
>> questino was asked to find out about applicability of those PDUs -
>> which protocols require which PDUs, and what is the motivation for
>> including each of those PDUs in draft-ietf-isis-layer2-05.txt.
>>
>>       - Ralph
>>
>>
>>       On May 5, 2010, at 8:21 AM 5/5/10, John E Drake wrote:
>>
>>
>>
>>               Ralph,
>>
>>               I think the point is that unlike TRILL, and IEEE
> 802.1aq,
>> OTV is proprietary - nobody other than its designers has any idea how
>> it works.
>>
>>               Publishing a few of its TLVs does absolutely nothing
>> regarding interoperability.
>>
>>               Thanks,
>>
>>               John
>>
>>
>>
>>                       -----Original Message-----
>>                       From: [e-mail] [mailto:isis-wg-
>> [e-mail]] On Behalf
>>                       Of Ralph Droms
>>                       Sent: Wednesday, May 05, 2010 5:07 AM
>>                       To: [e-mail]
>>                       Subject: Re: [Isis-wg] I-D
> Action:draft-ietf-isis-
>> layer2-05.txt
>>
>>                       Which of TRILL, IEEE 802.1ag and OTV use each of
> the
>> PDUs defined in
>>                       draft-ietf-isis-layer2-05.txt?  Is it possible
> to
>> devise an
>>                       applicability or usage table, similar to the
> table
>> for TLVs in section
>>                       5, for the new PDUs?
>>
>>                       - Ralph
>>
>>                       _______________________________________________
>>                       Isis-wg mailing list
>>                       [e-mail]
>>                       https://www.ietf.org/mailman/listinfo/isis-wg
>>
>>
>>
>>       _______________________________________________
>>       Isis-wg mailing list
>>       [e-mail]
>>       https://www.ietf.org/mailman/listinfo/isis-wg
>>
>>
>
>