Hi authors,

Please find below my detailed comments on
draft-ietf-pce-association-diversity. I originally started to review
-06. Thanks for posting -07 after Dhruv's comments, as it addressed some
on mine as well.

The main technical concern lies in section 4.6, in case no solution is
found by the PCE. Section 4.3, about SVEC, relies on PCRep with NO-PATH,
which is consistent with existing PCEP specification. IMHO, section 4.6
is confusing and incomplete. What about the following?

OLD
   [...] the PCE SHOULD
   reply with a PCUpd message containing an empty ERO.  In addition to
   the empty ERO Object, the PCE MAY add the NO-PATH-VECTOR TLV [...]

NEW
    [...] the PCE MUST
   reply to a request (PCEReq) with a PCRep message containing a NO-PATH
   object. In case of network event leading to an impossible strict
   disjointness, the PCE SHOULD send a PCUpd message containing an empty
   ERO to the corresponding PCCs. In addition to the NO-PATH or the
   empty ERO object, the PCE MAY add the NO-PATH-VECTOR TLV [...]

Complementary question: why an optional behavior (SHOULD) instead of
mandatory (MUST)?


Nits:
------
Global and usual nit: the flag name. The I-D has a collection of X
flag/X-flag/X-Flag/flag X/etc that needs consistency. Many PCEP
documents use "X flag".
------
Title
---
- s/communication Protocol (PCEP) extension for signaling LSP diversity
constraint/Communication Protocol (PCEP) Extension for LSP Diversity
Constraint Signaling/
------
Abstract
---
- s/Communication Protocol/communication Protocol/
- s/knows that LSPs in the same group/knows that the LSPs in the same group/
- s/needs to/need to/
------
2. Terminology
---
- s/Communication Protocol/communication Protocol/
------
3.  Motivation
---
- s/above, consider that/above, let us consider that/
- s/difficult. Whereas, computation/difficult, whereas computation/
- s/These messages uses/These messages use/
- s/the disjoint path computation/a disjoint path computation/
- s/disjoint group-ids/disjoint group IDs/
- s/should allow to overcome/allows to overcome/
- s/association source could/the association source could/
------
4. Protocol extension
---
- s/Protocol extension/Protocol Extension/
- s/Association group/Association Group/
- s/TLVs - Global Association Source or Extended Association ID are
included/TLVs - Global Association Source or Extended Association ID -
are included/
- s/to uniquely identifying/to uniquely identify/
- s/Association object -/Association object:/
- s/[I-D.ietf-pce-association-group]
specify/[I-D.ietf-pce-association-group] specifies/
- s/in the PCEP messages/in PCEP messages/
- s/Refer [I-D.ietf-pce-association-group]/Refer to
[I-D.ietf-pce-association-group]/
- s/more LSPs. But a PCE/more LSPs, but a PCE/
- s/in how many LSPs/in the number of LSPs/
- s/vendor specific behavioral information/vendor-specific behavioral
information/
- OLD
         When unset, PCE is allowed to relax disjointness
         by using either applying a requested objective function or any
         other behavior if no objective function is requested (e.g.:
         using a lower disjoint type (link instead of node) or relaxing
         disjointness constraint fully)
  NEW
         When unset, the PCE is allowed to relax disjointness
         by either applying a requested objective function (cf. section
         4.4 below) or using any other behavior if no objective function
         is requested (e.g. using a lower disjoint type (link instead of
         node) or fully relaxing disjointness constraint).

- s/The flags  L, N, and S/The L, N and S flags/
- s/The flag P/The P flag/
- s/the flag T/the T flag/
- s/both SVEC and ASSOCIATION object/both SVEC and ASSOCIATION objects/
- s/in SVEC object/in the SVEC object/
- s/with NO-PATH object/with a NO-PATH object/
- s/Disjointness objective functions/Disjointness Objective Functions/
- s/The PCEP OF-List TLV allow/Whereas the PCEP OF-List TLV allows/
- s/Incompatible OF codes/Incompatible OF code/
- s/listed below -/listed below:/
- The last example at the end of section 4.4 shows that the
specification doesn't prohibit redundant constraints. I would be nice to
add a sentence explicitly stating that in the previous paragraph.
- s/P-flag considerations/P Flag Considerations/
- s/fulfill the customer requirement/fulfill customer's requirements/
- s/Consider, this customer/Let us consider that this customer/
- s/( and/(and/
- s/allows a simple expression that/allows to simply express that/
- s/If PE->PE2/If PE1->PE2/
- Many P-flag/P-Flag to be fixed in section 4.5.
- s/Disjointness computation issues/Disjointness Computation Issues/
- s/T-bit/T flag/  [x2]
------
5. Security Considerations
---
- s/which do not/which does not/
- s/defines following new PCEP TLVs/defines the following new PCEP TLVs/
- s/PCEP-ERROR codes/PCEP-ERROR Codes/
- s/defines new Error-Type and Error-Value/defines new Error-Value
within existing Error-Type/
- s/Incompatible OF codes/Incompatible OF code/
------
8. Acknowledgements
---
- s/to author of/to the authors of/
------

Thanks,

Julien


_________________________________________________________________________________________________________________________

Ce message et ses pieces jointes peuvent contenir des informations 
confidentielles ou privilegiees et ne doivent donc
pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu ce 
message par erreur, veuillez le signaler
a l'expediteur et le detruire ainsi que les pieces jointes. Les messages 
electroniques etant susceptibles d'alteration,
Orange decline toute responsabilite si ce message a ete altere, deforme ou 
falsifie. Merci.

This message and its attachments may contain confidential or privileged 
information that may be protected by law;
they should not be distributed, used or copied without authorisation.
If you have received this email in error, please notify the sender and delete 
this message and its attachments.
As emails may be altered, Orange is not liable for messages that have been 
modified, changed or falsified.
Thank you.

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

Reply via email to