Re: [Ntp] WG LC for draft-ietf-ntp-packet-timestamps

Tal Mizrahi <tal.mizrahi.phd@gmail.com> Wed, 09 January 2019 06:39 UTC

MIME-Version: 1.0
References: <CA+RyBmWgQVVXv=ZntuNECjxgG-JdHCWD_ux9FKeR=9ihOOu-uw@mail.gmail.com>
In-Reply-To: <CA+RyBmWgQVVXv=ZntuNECjxgG-JdHCWD_ux9FKeR=9ihOOu-uw@mail.gmail.com>
From: Tal Mizrahi <tal.mizrahi.phd@gmail.com>
Date: Wed, 09 Jan 2019 08:39:25 +0200
Message-ID: <CABUE3X=kxKaRnj30=AsCJfAy=dY87NLRwK=GeF53Buxu2UW9mA@mail.gmail.com>
To: Greg Mirsky <gregimirsky@gmail.com>
Cc: draft-ietf-ntp-packet-timestamps@ietf.org, ntp@ietf.org
Content-Type: multipart/alternative; boundary="000000000000d17ccc057f00b9d5"
Archived-At: <https://mailarchive.ietf.org/arch/msg/ntp/0aywAhJ4BctynZAmgSI5lBXcyIs>
Subject: Re: [Ntp] WG LC for draft-ietf-ntp-packet-timestamps
Precedence: list

Hi Greg,

Many thanks for the comments.
We will address these comments in the next version of the draft.
More detailed responses below.

On Wed, Jan 9, 2019 at 5:53 AM Greg Mirsky <gregimirsky@gmail.com> wrote:

> Dear Authors, et al.,
> happy New Year!
> Apologies for the belated response.
> Firstly, I support the publication of this document and thank the authors
> for their work.
> Secondly, I have several comments to share, all are non-blocking and would
> appreciate you consider them as WG LC comments:
>
>    - "binary-encoded timestamp format" as the alternative to the clear
>    text format may be better than "packet timestamp" term to characterize the
>    subject of the document;
>
> We will clarify this point in the document.

>
>    - discussion of decisions, design choices that affect timestamp
>    accuracy and timestamp precision is very important but I don't see it in
>    the document. Perhaps, at this stage, references to the accuracy and the
>    resolution can be removed from the text to be discussed in another document
>    in the future;
>
> These aspects are mentioned in two places in the document: Section 5
(Synchronization Aspects) and Section 7 (Control Field).
Your point is well taken, and I will add more explicit language about the
fact that these aspects are expected to be discussed in future documents.

>
>    - regarding the number of timestamp formats supported by the new
>    protocol. I agree with your view that usually one format should be used.
>    But there could be scenarios that systems have different wall clocks, i.e.,
>    their native time format. In such cases, it is advantageous to enable
>    collection of the timestamp in the native format, NTP or PTP. I think that
>    the document should mention that multiple binary-encoded timestamp formats
>    may be used;
>
> Agreed.

>
>    - the table in section 6 presents RFC 8186 as supporting only PTP
>    Truncated format for timestamps in TWAMP Test packet. Actually, with the
>    RFC 8186 TWAMP Session-Sender and Session-Reflector may use either NTP
>    64-bit or PTP (capability advertisement and negotiation in TWAMP-Control).
>
> Right. Will update the document.

>
> Regards,
> Greg
>


Regards,
Tal.

Re: [Ntp] WG LC for draft-ietf-ntp-packet-timesta… Tal Mizrahi