Re: [ippm] Document Action: 'Test Plan and Results for Advancing RFC 2680 on the Standards Track' to Informational RFC (draft-ietf-ippm-testplan-rfc2680-05.txt)

Hi Nalini,

You raise an important point in question 1.
The original authors of 2679 (delay, or 2680 for loss) did not include
the constant waiting time threshold as an input parameter
(and so they didn't spend much time describing it).
However, it is included explicitly  
(as parameter Tmax) in https://tools.ietf.org/html/rfc6049#section-3.1.1
and discussed in more detail in https://tools.ietf.org/html/rfc6703#section-4.1
Also, ITU-T Rec Y.1540 for IP metrics uses the parameter Tmax as
part of their loss & delay definitions, and we have a long-standing goal
to move IPPM and ITU-T definitions closer together when possible
(rather than highlight the differences and leave them alone).

In short, I think it would be a very reasonable clarification (for both "bis" drafts)
to make this constant waiting time parameter more explicit now - excellent catch!

on 2, wire time error, there would continue to be some
small time difference between external appearance of a packet
and the interface hardware time stamp, although very much less
difference than TS up in the stack. So it's a smaller error
but I think the discussion still holds.

thanks for your time and careful reading!
Al
________________________________________
From: ippm [ippm-bounces@ietf.org] On Behalf Of Nalini Elkins [nalini.elkins@insidethestack.com]
Sent: Monday, May 12, 2014 11:09 AM
To: ippm@ietf.org
Subject: [ippm] Document Action: 'Test Plan and Results for Advancing RFC 2680 on the Standards Track' to Informational RFC (draft-ietf-ippm-testplan-rfc2680-05.txt)

Al,

Some comments on draft-morton-ippm-2679-bis-04.txt:

1. The items in section 1 of RFC2679bis (delay) appear to be implemented in the draft but I am having a bit of trouble with understanding:

"Further, enforcing a specific constant waiting time on stored singletons of one-way delay is compliant with this specification and may allow the results to serve more than one reporting audience."

I am not finding the discussion of constant waiting time with regard to singletons.  Is it in the test plan document?

2.  I also had a comment on Section 3.7.2 Errors or uncertainties related to Wire-time vs Host-time

I have had some discussions with folks about the idea of offloading the processing of the timestamps onto hardware.  That is, performing these kind of functions on a chip.  I wonder if that changes the conversation here a bit?

Thanks,

Nalini Elkins
Inside Products, Inc.
(831) 659-8360
www.insidethestack.com

_______________________________________________
ippm mailing list
ippm@ietf.org
https://www.ietf.org/mailman/listinfo/ippm