Randy
I would suggest some wording if I knew what was intended but as yet, I don't:-(.
I suspect that Bill's description - use the next available integer in sequence -
may be what is intended but, for me, that is not the sense of the words. Off
list:-(, I did get a different interpretation - from one who was involved in
the earlier discussion of monotonic - that any index value would do as long as
the order of the entries in the table matched the order of the hops. So I still
think that there is a minor ambiguity here
Tom Petch
----- Original Message -----
From: "Randy Presuhn" <randy_presuhn(_at_)mindspring(_dot_)com>
To: "Bill Fenner" <fenner(_at_)gmail(_dot_)com>;
<j(_dot_)schoenwaelder(_at_)iu-bremen(_dot_)de>;
"Tom.Petch" <sisyphus(_at_)dial(_dot_)pipex(_dot_)com>; "Bill Strahm"
<bill(_at_)strahm(_dot_)net>; "iesg"
<iesg(_at_)ietf(_dot_)org>; "ietf" <ietf(_at_)ietf(_dot_)org>
Cc: "Disman" <disman(_at_)ietf(_dot_)org>
Sent: Tuesday, February 28, 2006 10:16 PM
Subject: Re: Last Call: 'Definitions of Managed Objects for Remote
Ping,Traceroute, and Lookup Operations' to Proposed Standard
Hi -
If the document gives a false impression that the values of
traceRouteHopsHopIndex could be interpreted as hop numbers,
an editorial change to dispel that notion would make sense.
(Likewise, if "consecutive integers starting at one" was the intent, and
is what current implementations actually do, then we should say so.)
I can see how the last two sentences of the last paragraph of
the DESCRIPTION might lead to such a reading. Does
someone have some replacement text they'd like to propose
to make things clearer?
Randy, disman chair
----- Original Message -----
From: "Bill Fenner" <fenner(_at_)gmail(_dot_)com>
To: <j(_dot_)schoenwaelder(_at_)iu-bremen(_dot_)de>; "Tom.Petch"
<sisyphus(_at_)dial(_dot_)pipex(_dot_)com>;
"Bill Strahm" <bill(_at_)strahm(_dot_)net>; "iesg" <iesg(_at_)ietf(_dot_)org>;
"ietf" <ietf(_at_)ietf(_dot_)org>
Sent: Monday, February 27, 2006 10:48 PM
Subject: Re: Last Call: 'Definitions of Managed Objects for Remote
Ping,Traceroute, and Lookup Operations' to Proposed Standard
Juergen,
I assumed, from reading in traceRouteHopsHopIndex about the behavior
when a path changes, that the only safe thing for a manager to do is
to read the hops from the table and render them to the user in order
of increasing traceRouteHopsHopIndex but without necessarily showing
the traceRouteHopsHopIndex to the user -- that it was perfectly
reasonable for hops 1,2,3,4 of a 4-hop path to be numbered 1,8,12,35
(assuming that they started 1,2,3,4 but there were lots of path
changes during the test).
I think some people are assuming that the intention was that the
values should be 1,2,3,4 (i.e., HopIndex == hop number) and that's why
they're asking for a different definition. Perhaps the right
direction could be to clarify that there is no connection between the
value of HopIndex and traceroute hop, other than the ordering.
Bill
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf