Hi Doug, all,

I am also in line with Silvana and Stefano.

I would be also interested to better understand the use case that is addressed, 
and what makes it so specific that it requires the development of a dedicated 
PTP profile.
By nature, enterprise networks have always been very close to telecom networks 
(I might however have missed an important point).

Indeed, in my opinion, most of the PTP options that are included in this draft 
are more or less available in PTP profiles already developed or under 
development. Just to give an idea:


-          The mixed unicast/multicast arrangement, if required, is something 
that was originally planned in G.8265.1 (see Appendix I). This is something 
which could be finalized if a valid use case shows the interest in this mode 
(ITU stopped at some point working on this mode to focus on the full unicast 
mode, but there was originally interest from various companies). As mentioned 
by Stefano, this arrangement seems only relevant to frequency synchronization, 
and not accurate time synchronization, due to the asymmetry generated.


-          The possibility to include PTP unaware nodes in the chain 
corresponds to the so-called "partial support" case, for which a study item has 
been started in ITU-T. This may result in the definition of a dedicated PTP 
profile, currently envisaged in future G.8275.2.

As you can see, there is some overlapping with on-going studies outside TICTOC, 
so it might be worth trying to coordinate to avoid that similar work would be 
developed twice.

Finally, I also share the warnings that defining a PTP profile only (protocol 
aspects) cannot ensure that the relevant performance will be met.

I won't be in Orlando to further discuss this work, but I'll try to follow 
remotely what happens, maybe with the help of some colleagues, and provide 
comments if useful.

Thanks.

BR,

Sébastien

De : [email protected] [mailto:[email protected]] De la part de 
Rodrigues, Silvana
Envoyé : mardi 5 mars 2013 04:07
À : Stefano Ruffini; Doug Arnold; [email protected]
Objet : Re: [TICTOC] PTP Enterprise Profile draft

Doug,

Thanks for the draft. I also agree with Stefano's comment and I have a few 
additional comments.

Item 3 and 4 from Stefano's email below, I think that there is a need to add 
some warning in the text, as the way it is written, it does look like that  if 
this profile is implemented over any network the 1us requirement will be met.

Section 10, it states that slave may use an Acceptable Master list. How this 
will work when transparent clocks are used and the source address in the TCs 
are changed to the TC source address? The Slave will no longer have the source 
address of the masters.

Section 12, it states that Unicast negotiation is forbidden. How the packet 
rate will be negotiated between the master and slave for Unicast Mode?

Thanks,

Silvana


From: [email protected] [mailto:[email protected]] On Behalf Of 
Stefano Ruffini
Sent: Monday, March 04, 2013 11:20 AM
To: Doug Arnold; [email protected]
Subject: Re: [TICTOC] PTP Enterprise Profile draft

Hi Doug,


thanks for your draft .


There are a few points that I think would need some clarification.

1) As also mentioned at last tictoc meeting, there is a risk of multiplication 
of profiles.
It is not fully clear in the Introduction what is so specific to enterprise 
networks to requiring the definition of a specific profile.



2) the proposed profile would support a combinations of various options (e.g. 
various modes defined in section 6) but the task of a profile should normally 
also be to minimize the number of IEEE1588 options. Are all these options 
really required ?


3) section 4 indicates a target requirement of 1 us : is there a reference for 
that?

4) It should be noted that the performance (e.g. 1 us) can not be guaranteed by 
a profile as such.
Other aspects, such as network reference model, clock characteristics, also 
have to be defined in order to provide some guarantee on the achievable 
performance.
How are these planned to be addressed? Will it make reference to ITU-T 
recommendations?


5) Concerning the "Hybrid" and "Unicast" mode in section 6.3 (see also remark 
from Laurent on the confusing terminology used), in case of "partial support" , 
the potential risk of adiditonal asymmetries should be mentioned (see also 
Appendix I in G.8265.1) .
This is not an issue in case of frequency sync but it is in case of time sync.


Best Regards
Stefano

________________________________
From: [email protected]<mailto:[email protected]> 
[mailto:[email protected]] On Behalf Of Doug Arnold
Sent: lunedì 25 febbraio 2013 22.52
To: [email protected]<mailto:[email protected]>
Subject: [TICTOC] PTP Enterprise Profile draft
Here is a first cut at a draft of a PTP profile foe enterprise networks.  Any 
feedback/suggestions would be much appreciated.

I enclose both a MS Word version and a pure text version, since I used the Joe 
Touch's template.

//Doug


_________________________________________________________________________________________________________________________

Ce message et ses pieces jointes peuvent contenir des informations 
confidentielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu ce 
message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages 
electroniques etant susceptibles d'alteration,
France Telecom - Orange decline toute responsabilite si ce message a ete 
altere, deforme ou falsifie. Merci.

This message and its attachments may contain confidential or privileged 
information that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and delete 
this message and its attachments.
As emails may be altered, France Telecom - Orange is not liable for messages 
that have been modified, changed or falsified.
Thank you.

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

Reply via email to