Re: [Isis-wg] *** Revised ID needed *** Last Call: draft-ietf-mpls-number-0-bw-te-lsps (ALink-Type sub-TLV to convey the number of Traffic EngineeringLabel Switched Paths signalled with zero reserved bandwidthacross a link) to Propose

On 8/19/08 7:47 PM, "JP Vasseur" <jvasseur@cisco.com> wrote:

> Dear WGs,
> 
> We just posted the new revision of draft-ietf-mpls-number-0-bw-te-lsps:
> http://www.ietf.org/internet-drafts/draft-ietf-mpls-number-0-bw-te-lsps-11.txt
> 
> The changes takes into account the comment received during WG LCs and are
> summarized below:
>>> 
>>>>>  
>>>>>  Here are some comments from various lists that I can find on the ISIS
>>>>> aspects:
>>>>>  
>>>>>  
>>>>>  -----------------------------------
>>>>>  "Section 1:
>>>>>  
>>>>>  s/Constraint/Constrained
>>>>>  
>>>>>  Section 2 Paragraph 3 First Sentence
>>>>>  
>>>>>  s/assumption/assumptions
>>>>>  
>>>>>  s/unconstrained TE Label Switched Path/unconstrained TE Label Switched
>>>>>  Paths
>>>>>  (plural "paths")
>>>>>  
>>>>>  JP> Fixed.
>>>>>  
>>>>>  Section 3
>>>>>  
>>>>>  As the section is discussing two different sub-TLVs (one for IS-IS and
>>>>>  one for OSPF) the first sentence should read:
>>>>>  
>>>>>  "Two Unconstrained TE LSP count sub-TLVs are defined that specify the
>>>>>     number of TE LSPs signalled with zero bandwidth across a link.
>>>>>  
>>>>>  JP> Fixed.
>>>>>  
>>>>>  Section 3.1
>>>>>  
>>>>>  It might be worth mentioning that the new sub-TLV could also appear in
>>>>>  the MT IS-Neighbor TLV (222) - but I won't insist on it as it is
>>>>>  generally assumed that anything that appears in TLV 22 could also appear
>>>>>  in TLV 222.
>>>>>  "
>>>>>  JP> OK I added a reference.
>>>>>  ---------------------------------------
>>>>>  
>>>>>  "
>>>>>  if a link flaps and a substantial number of tunnels are going through
>>>>>  it, these tunnels will be re-routed through other links. This will
>>>>>  trigger flooding of isis lsp's in order to advertise/update the
>>>>>  te-lsp-count subtlv, right ? Do we need to specify anything in order to
>>>>>  prevent storms ?
>>>>>  "
>>>>>  
>>>>>  This last one is relevant to be addressed w/ some modified text in the
>>>>> draft. More to come as I get it.
>>>>>  
>>>>>  JP> Which is no different than for any other TE-related TLVs. Even with
>>>>> non-0 bw TE LSPs, if a link flaps they will get rerouted on other links,
>>>>> which will change the reserved bandwidth and will trigger the flooding of
>>>>> ISIS LSP or OSPF LSA to reflect the updated reserved bandwidth ? This is
>>>>> why we added ³Similarly to other MPLS Traffic Engineering link
>>>>> characteristics, LSA/LSP origination trigger mechanisms are outside the
>>>>> scope of this document.²
>>>>>  
>>>>>  That being said, I added the following : ³care must be given to not
>>>>> trigger the systematic flooding of a new IS-IS LSP or OSPF LSA with a too
>>>>> high granularity in case of change of the number of unconstrained TE
>>>>> LSPs.²
>>>>>  
>>>>>  JP> Furthermore, I checked IANA actions, all correct.
>>>>>  
>>>>>  -Dward
>>>>> 
>>>>> 
>>>>> Thanks.
>>>>> 
>>>>> JP.
>>>>>  
>>>>>  
>>>>>  
>  