[Teas] RtgDir Review of draft-ietf-teas-mpls-tp-rsvpte-ext-associated-lsp-02

["[e-mail]" <[e-mail]>]

Hello
 
I have been selected as the Routing Directorate reviewer for this draft.
The Routing Directorate seeks to review all routing or routing-related
drafts as they pass through IETF last call and IESG review, and
sometimes on special request. The purpose of the review is to provide
assistance to the Routing ADs. For more information about the Routing
Directorate, please see ​
http://trac.tools.ietf.org/area/rtg/trac/wiki/RtgDir
 
Although these comments are primarily for the use of the Routing ADs, it
would be helpful if you could consider them along with any other IETF
Last Call comments that you receive, and strive to resolve them through
discussion or by updating the draft.
 
Document: draft-ietf-teas-mpls-tp-rsvpte-ext-associated-lsp-02
Reviewer: Lizhong Jin
Review Date: 19 February 2015
IETF LC End Date: N/A
Intended Status: Standard Track
 
Summary:
The document is basically ready for publication. I found some minor issues, and hope
to see the clarification.

Commnets:
I paticipate the initial discussion of this draft. I am lucky to review it again. Overall, the
processing rule of independent provisioning hope to be explicitly described. 
No major issues, some minor issues need to be clarified.

Major issues: No

Minor issues:
Section 3.2
In each of the situations described above, both provisioning models 
are applicable. 
[Lizhong]: for the second situations above, how could you let the reverse LSP 
to be existed before the forward LSP for single side provisioning? Wouldn't the 
reverse LSP trigger another reverse LSP?

Section 3.2.1. 
For the single sided provisioning model, creation of reverse LSP1 is 
triggered by LSP2 or creation of reverse LSP2 is triggered by LSP1. 
When creation of reverse LSP2 is triggered by LSP1, LSP1 is 
provisioned first (or refreshed if LSP1 already exists) at node A. 
[Lizhong]: LSP1 and LSP2 is in Firgure1? Better to explicitly say that in
the document.

 A similar procedure is used if LSP2 is provisioned first at node B 
and the creation of reverse LSP1 at node A is either triggered by 
LSP2 or the reverse LSP1 existed. In all three scenarios, the two 
unidirectional LSPs are bound together to form an associated 
bidirectional LSP based on identical (Extended) ASSOCIATION Objects 
in the two LSPs' Path messages. 
[Lizhong]: I doubt if the following scenario is realistic in Single Sided Provisioning. 
LSP2 is provisioned first, reverse LSP1 at node A is existed before LSP2. 
The what is the Association Type in reverse LSP1? Before LSP2, will the reverse 
LSP1 trigger another LSP? 

Section 5. Processing Rules 
In general, the processing rules for the ASSOCIATION Object are as 
specified in [RFC4872] and Extended ASSOCIATION Object are specified 
in [RFC6780]. Following sections describe the rules for processing 
(Extended) ASSOCIATION and REVERSE_LSP objects for associated 
bidirectional LSPs. 
[Lizhong]: across the draft, it is not explicitly saying what is the processing 
rules for independent provisioning. It is better to say it here or other place.

Section 5.1
(Extended) ASSOCIATION Objects with both single sided and double sided 
Association Types MUST NOT be added in the same Path message. 
[Lizhong]: what if two types exist together? Only use the first one?

Section 5.2
The REVERSE_LSP Object MUST NOT be included in a REVERSE_LSP Object. 
[Lizhong]: typo here?

Section 5.3
In particular, any object that was copied as part of initial Path message creation MUST
be copied when modified. 
[Lizhong]: not understood "copied when modified", is it "copied after modified"?

In both cases, when the egress node receives a PathTear message the 
node MUST remove the associated reverse LSP using Standard PathTear 
message processing. Tear down of the reverse LSP for other reasons 
SHOULD NOT trigger removal of the initiating LSP, but SHOULD result 
in the egress node sending a PathErr with Error code "Admission 
Control Failure (01) [RFC2205]" and Sub-code "Reverse LSP Failure" 
defined in this document. 
[Lizhong]: the above description is not accurate. What if the egress node have
forward LSP down because of local link down? In that case, it will not receive
PathTear, it is still need to tear down the reverse LSP? It should say, whenever 
the forward LSP is down, the reverse LSP SHOULD be removed. 

Regards
Lizhong