Acee –

From: Acee Lindem (acee)
Sent: Monday, November 27, 2017 7:16 PM
To: Les Ginsberg (ginsberg) <[email protected]>; Naiming Shen (naiming) 
<[email protected]>
Cc: Christian Hopps <[email protected]>; [email protected]; [email protected]
Subject: Re: [Isis-wg] WG Last Call for draft-ietf-isis-reverse-metric-07

Hi Les,

I guess we know the origin of the term “metric-offset” ;^)

[Les:] Always good to know who to blame. ☺

From: "Les Ginsberg (ginsberg)" <[email protected]<mailto:[email protected]>>
Date: Monday, November 27, 2017 at 6:53 PM
To: Acee Lindem <[email protected]<mailto:[email protected]>>, "Naiming Shen 
(naiming)" <[email protected]<mailto:[email protected]>>
Cc: Christian Hopps <[email protected]<mailto:[email protected]>>, 
"[email protected]<mailto:[email protected]>" 
<[email protected]<mailto:[email protected]>>, 
"[email protected]<mailto:[email protected]>" 
<[email protected]<mailto:[email protected]>>
Subject: RE: [Isis-wg] WG Last Call for draft-ietf-isis-reverse-metric-07

Acee –

The IS-IS draft is more flexible in this regard than the OSPF equivalent 
(https://tools.ietf.org/html/draft-ietf-ospf-link-overload-10).

In the IS-IS draft the text states:

“The Metric Offset field contains a 24-bit unsigned integer of an IS-
   IS metric that a neighbor SHOULD add to the existing, configured
   "default metric" contained within its IS Neighbors TLV…”

This allows that the operator could choose to set the neighbor metric to 
something other than “max-metric-1”.


Contrast this with the OSPF Draft which states:

“The node that has the link to be taken out of service MUST set metric
   of the link to MaxLinkMetric (0xffff)…”

Why are you comparing this draft to OSPF Link Overload? If you know the metric 
value, you don’t need any metric field (as is the case for OSPF Link Overload).

The text in the IS-IS draft needs to remain as it is.

The argument for keeping the text as is is flawed. Using "metric” or “reverse 
metric” for the identifier of the field would simply align it with the 
identifier for the advertising TLV. In the context of a reverse metric (the 
advertised TLV), it is not a metric offset even though it is added to the 
existing metric in the SPF computation.

[Les:] The draft defines the usage of the metric value advertised in the 
Reverse Metric TLV as a value that

“a neighbor SHOULD add to the existing, configured  default metric”

This definition was the same even in earlier versions (V5 and earlier) where 
the name of the field was “metric” rather than “metric offset”.

I always thought the purpose of defining it as an offset was to prevent an 
operator from accidentally making the metric smaller than the value configured 
on the neighbor. You seem to be arguing that this is the wrong way to do things 
– that you should always send an absolute value – but this does mean there is 
some risk that the value the neighbor advertised upon receipt of the Reverse 
Metric TLV could be smaller than it currently is. This could be prevented by 
normative language in the draft to say the neighbor should ignore such cases.

The authors should comment on whether they would prefer to move to your 
proposal. Personally I see no need to change the draft.
But, in terms of the best name for the field in the TLV I think it should 
represent how that value is used. Currently that is as an “offset”. If you 
convince the authors to change things and make the value an absolute then I 
would agree the term “metric” would be better. All I want is clarity and 
consistency.

    Les

Thanks,
Acee

   Les


From: Isis-wg [mailto:[email protected]] On Behalf Of Acee Lindem (acee)
Sent: Monday, November 27, 2017 3:33 PM
To: Naiming Shen (naiming) <[email protected]<mailto:[email protected]>>
Cc: Christian Hopps <[email protected]<mailto:[email protected]>>; 
[email protected]<mailto:[email protected]>; 
[email protected]<mailto:[email protected]>
Subject: Re: [Isis-wg] WG Last Call for draft-ietf-isis-reverse-metric-07

Hi Naiming,

From: "Naiming Shen (naiming)" <[email protected]<mailto:[email protected]>>
Date: Monday, November 27, 2017 at 4:38 PM
To: Acee Lindem <[email protected]<mailto:[email protected]>>
Cc: Christian Hopps <[email protected]<mailto:[email protected]>>, 
"[email protected]<mailto:[email protected]>" 
<[email protected]<mailto:[email protected]>>, 
"[email protected]<mailto:[email protected]>" 
<[email protected]<mailto:[email protected]>>
Subject: Re: [Isis-wg] WG Last Call for draft-ietf-isis-reverse-metric-07


Hi Acee,

thanks for the comments. replies inline with <NS>…</NS>

On Nov 24, 2017, at 3:05 PM, Acee Lindem (acee) 
<[email protected]<mailto:[email protected]>> wrote:

Hi,

I support publication of the subject document. I do have comments.

  1. Why do you call the reverse-metricn field, “Metric Offset”? This
sounds like a remnant of some bad 20th century CLI….

<NS>
the draft used to have the field as just ‘Metric’, since version 06, we 
responded
to the comments:


Section 2

From the description what is being advertised in the new TLV is not a metric 
but a metric offset i.e. you want the receiving IS to add the advertised value 
to its existing configured metric. Identifying the metric field as "metric 
offset" would make this point more clearly.


which is a good point I think. do you think there is an alternative name we
can use in replacing the ‘offset’ here?
</NS>

How about just calling it the “Reverse Metric”?

Thanks,
Acee





  2. Please split the single sentence description immediately following
figure 1 into multiple sentences. Maybe just refer to “Elements of
Procedure” sections rather than one incomprehensible sentence.
  3. Pervasive editorial comment, the plural of acronyms is does NOT have
an apostrophe. For example, it is TLVs, not TLV’s.

<NS>
will change those.

thanks.
- Naiming
</NS>




Thanks,
Acee


On 11/15/17, 5:43 PM, "Isis-wg on behalf of Christian Hopps"
<[email protected]<mailto:[email protected]> on behalf of 
[email protected]<mailto:[email protected]>> wrote:




The authors have asked for and we are starting a WG Last Call on

https://datatracker.ietf.org/doc/draft-ietf-isis-reverse-metric/

which will last an extended 3 weeks to allow for IETF100.

Thanks,
Chris.

_______________________________________________
Isis-wg mailing list
[email protected]<mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/isis-wg

_______________________________________________
Isis-wg mailing list
[email protected]<mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/isis-wg

_______________________________________________
Isis-wg mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/isis-wg

Reply via email to