Dear Yaakov,
perhaps my wording wasn't precise. I think of ACH option not as
PTP-over-MPLT-TP transport but as PTP interworking between 1588
MPLS-TP domain and 1588 non-MPLS-TP (IP/UDP) domain.

Regards,
Greg

On Wed, Jul 14, 2010 at 10:43 AM, Yaakov Stein <[email protected]> wrote:
> Greg
>
> Yes, a dedicated Ach is a possibility. I mentioned that as using the GAch 
> format. I actually like this idea.
>
> But why do you believe that this is any less of a layer violation ?
> The timestamps come from layer 1 (they are physical mechanisms)
> while MPLS is somewhere around layer 2 or higher.
> If the MPLS is carried over Ethernet or some other CRC protected layer 2, the 
> CRC would need to be updated.
> If there is some integrity protection at layer 1 or 2 then it will break when 
> we update the TC field,
> and the packet would be discarded.
> If there is layer 2 encryption, the mechanism won't be applicable unless we 
> open and close the encryption
> at a link-by-link basis.
>
> Y(J)S
>
> -----Original Message-----
> From: Greg Mirsky [mailto:[email protected]]
> Sent: Wednesday, July 14, 2010 19:46
> To: Yaakov Stein
> Cc: Alexander Vainshtein; [email protected]
> Subject: Re: [TICTOC] FW: 1588 over MPLS draft
>
> Dear Yaakov and All,
>
> <snip>
>
>>
>> <YJS>  YES !  This is my method number 3.  A new reserved label. Of course 
>> we need to convince the MPLS
>> WG to give us one (there aren't that many ...)
>>
> GIM>> What if instead of LSP/PW as PTP transport we use dedicated to
> 1588 ACH in an MPLS-TP DCN? TC-capable nodes might be interconnected
> by DCN links and processing TC would not violate network layers.
>
> Regards,
> Greg
>
_______________________________________________
TICTOC mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tictoc

Reply via email to