Re: [mpls] [PWE3] [mpls-tp] FW: Changes to PW ACH Channel Typeallocationpolicy

[Alexander Vainshtein <Alexander.Vainshtein at ecitele.com>]

Maarten and all,
Please see some comments inline below. The unrelated text has been snipped to improve readability.

Regards,
        Sasha
> -----Original Message-----
> From: pwe3-bounces at ietf.org [mailto:pwe3-bounces at ietf.org] On
> Behalf Of Maarten Vissers
> Sent: Monday, October 06, 2008 9:49 AM
> To: 'Shahram Davari'; 'Thomas D. Nadeau'; 'Shane Amante'
> Cc: l2vpn at ietf.org; mpls at ietf.org; 'BOCCI Matthew';
> ccamp at ietf.org; mpls-tp at ietf.org; pwe3 at ietf.org; 'BUSI ITALO'
> Subject: Re: [PWE3] [mpls-tp] FW: Changes to PW ACH Channel Type allocation policy
>
> Sharam,
>
> I agree.
>
> In my understanding the PW ACH Channel Type provides the same
> function as the Ethertype in Ethernet networks. IEEE provides a means to buy an
> Ethertype codepoint (http://standards.ieee.org/regauth/ethertype/eth.txt).
> The requesting organization can decide to keep their associated application
> private ("Protocol unavailable.") or to make it public.
>
> As far as I understand this Ethertype allocation policy is an industry
> standard and could be copied for PW ACH Channel Type allocation policy.
>
[Sasha] IMHO and FWIW your description of the IEEE process is somewhat inaccurate.
I am not an expert in the field, and Shahram (and others) will probably provide a much more accurate description
of the process, but still:

1. As you've mentioned, IEEE *SELLS* its Ethertype codepoints - at USD $2,500 per a *PUBLIC* codepoint.
    AFAIK, the IETF does not charge anybody  for its codepoints (hopefully this will not result in any bright ideas:-)

2. There is no such thing as a *PRIVATE* Ethertype codepoint at IEEE  (at least, I did not find any references
    on their Web site). AFAIK, the only private values you can obtain from IEEE are OUIs,
    and you must pay USD $1,650/year per private OUI to maintain privacy (that on top of USD $2,000 per a public OUI)

3. IEEE still requires description of the protocol before it approves your registration request, and it takes at least 90
    days before a new code point is granted. (From my experience when dealing with other SDOs IEEE requires an approved
    protocol spec lacking just the new Ethertype value in order to grant a new Ethertype codepoint.)

On top of all that, the most important issue here is that ACH has never been intended to be a new multiprotocol network layer.
This point has been raised already by the others.

Unfortunately your analogy between the ACH type and Ethertype only strengthens the suspicions that this can easily happen with an over-relaxed codepoint allocation policy.

... snipped to the end...

_______________________________________________
mpls mailing list
mpls at ietf.org
https://www.ietf.org/mailman/listinfo/mpls