Hi Stewart, 

looking or modifying anything beyond the top label is technically LV in MPLS. 
The only difference your approach has is that it doesn't need the offset to 
update the CF field, but the drawback is that it requires new protocol.

Note that the offset provided in Signaling is to the start of PTP PDU after all 
IP, UDP or Ethernet encapsulation and doesn't need to track it back to find out 
what is the encapsulation.


Thx
Shahram


Thx
Shahram

----- Original Message -----
From: Stewart Bryant [mailto:[email protected]]
Sent: Sunday, November 27, 2011 02:17 AM
To: Shahram Davari
Cc: '[email protected]' <[email protected]>; '[email protected]' <[email protected]>
Subject: Re: [TICTOC] FW: I-D Action: draft-ietf-tictoc-1588overmpls-02.txt

On 27/11/2011 02:51, Shahram Davari wrote:
> Hi Stewart,
>
> Your proposal also does layer violation.
Not in a transit time recording FEC. Time recording is a property of the 
FEC.

> TC by nature does layer violation. Advantage of our approach is that it uses 
> existing PW encapsulation, and doesn't require new protocol.
That depends on how it is done.

Recording the transit time and feeding that time along with the
packet into a BC is not LVing.

Also this draft is not just about PW encaps.
>
> Your solution can be further explored in a new draft as a more generic 
> solution, noting that requires new protocol and new parsing.
>
Compared to existing P routers you require new protocol and new parsing.

Stewart


_______________________________________________
TICTOC mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tictoc

Reply via email to