For NTP, It is not required to use this draft for transporting it over MPLS. 
Since NTP has not yet defined Transparent Clock, and  NTP accuracy as of now is 
low and does not require time stamping  in Hardware.

Regards,
Shahram


On Jun 7, 2013, at 9:08 AM, "Gregory Mirsky" 
<[email protected]<mailto:[email protected]>> wrote:

Hi Shahram,
and hence are my questions:

  *
Is proposed mechanism required to transport other timing protocols over MPLS 
network
  *
Is use of PTP LSP to transport NTP messages justified or not

    Regards,
        Greg

________________________________
From: Shahram Davari [mailto:[email protected]]
Sent: Friday, June 07, 2013 9:01 AM
To: Gregory Mirsky
Cc: Yaakov Stein; Amit Oren; Bhatia, Manav (Manav); Roberts, Peter (Peter); 
[email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Subject: Re: NTP in draft-ietf-tictoc-1588overmpls-04

Greg,

This draft was originally only for PTP, however our Area Director asked the 
draft to be made more generic so that it can apply to any timing protocol.

Regards,
Shahram


On Jun 7, 2013, at 8:28 AM, "Gregory Mirsky" 
<[email protected]<mailto:[email protected]>> wrote:

Yaakov,
I'm rather sceptical of using "If you build it, he will come" as protocol 
development guiding principle. Personally I prefer KISS.

    Regards,
        Greg

________________________________
From: Yaakov Stein [mailto:[email protected]]
Sent: Friday, June 07, 2013 7:55 AM
To: Gregory Mirsky; Shahram Davari; 
[email protected]<mailto:[email protected]>; Bhatia, Manav (Manav); Roberts, 
Peter (Peter); [email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Subject: RE: NTP in draft-ietf-tictoc-1588overmpls-04

Greg
This version makes the mechanism more general so that in future other 
time-sensitive protocols
(NTP, delay measurement OAM, generic header for MPLS) could avail themselves of 
its timestamping/correction features.
It is true that the only real case so far is 1588.
Y(J)S
From: [email protected]<mailto:[email protected]> 
[mailto:[email protected]] On Behalf Of Gregory Mirsky
Sent: 06 June, 2013 21:23
To: Shahram Davari; [email protected]<mailto:[email protected]>; Bhatia, 
Manav (Manav); Roberts, Peter (Peter); 
[email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Subject: [TICTOC] NTP in draft-ietf-tictoc-1588overmpls-04
Dear Authors, et al.,
The draft-ietf-tictoc-1588overmpls-04 is currently titled as "Transporting 
Timing messages over MPLS Networks" and states that both timing synchronization 
protocols, PTP and NTP, require use of PTP LSP to transport timing messages 
over an MPLS network. After reading RFC 5905 I am not sure that PTP LSP, with 
its constraints and additional complexity, is required to transport NTPv4 
messages over an MPLS network. And first paragraph in Section 3 Problem 
Statement of draft-ietf-tictoc-1588overmpls-04 explains motivation for the PTP 
as
"There is a need to transport Timing messages over MPLS networks while 
supporting the Transparent Clock (TC), Boundary Clock (BC) and Ordinary Clock 
(OC) functionality in the LER and LSRs in the MPLS network."
but says nothing of a kind about NTP.
        Regards,
                Greg
_______________________________________________
TICTOC mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tictoc

Reply via email to