Hierarchical Stateful Path Computation Element (PCE)

Note: This ballot was opened for revision 13 and is now closed.

(Deborah Brungard) Yes

(Alissa Cooper) No Objection

Roman Danyliw (was Discuss) No Objection

Comment (2019-10-15 for -14)
No email
send info
Thanks for addressing my DISCUSS point.

Benjamin Kaduk (was Discuss) No Objection

Comment (2019-10-18 for -14)
Thank you for addressing my Discuss point!

I would consider including the conclusion from our discussion about what
would happen if a PCEP speaker does not support stateful H-PCE but the peer
assumes it does, perhaps in an operational considerations section, but this does
not rise to a Discuss-level point.  For convenience, this was discussed as:

% I further did a mental exercise for PCC -> C-PCE -> P-PCE and assumed               
% all support stateful and H-PCE extension but what happens when any                  
% PCEP speaker does not support stateful H-PCE but the peer assumes that              
% it does. On further PCEP message exchange, the messages may not get                 
% further propagated and thus at worse would not lead to the stateful                 
% H-PCE based 'parent' control of the LSP. This is something any peer                 
% should be prepared for anyways.

(Suresh Krishnan) No Objection

(Mirja Kühlewind) No Objection

(Barry Leiba) No Objection

Comment (2019-09-14 for -13)
Thanks for this document.  I have only editorial suggestions (albeit, a lot of them).  There's no need to reply in any detail; just please consider adopting these suggestions.  Thanks.

— Abstract —

   computing new traffic engineered LSPs, and for associated and

Need a hyphen in “traffic-engineered”.

— Section 1.1 —

   This document presents general considerations for stateful PCEs, and
   not Stateless PCEs, in the hierarchical PCE architecture.  In
   particular, the behavior changes and additions to the existing
   stateful PCE mechanisms (including PCE-initiated LSP setup and active
   PCE usage) in the context of networks using the H-PCE architecture.

The second “sentence” is not a complete sentence.  One way to fix that is by changing “In particular,” to “It focuses on”.

   stitching Per Domain LSPs.

Throughout, “Per-Domain” needs a hyphen.

— Section 1.2 —

In the title, “Use Cases” should *not* be hyphenated (it’s a noun, and not being used as a modifier).  There’s also an instance of this in Section 3.

   As per [RFC6805], in the hierarchical PCE architecture, a P-PCE

The second comma doesn’t belong.

   It could also be a change in topology at the P-PCE
   such as inter-domain link status change.

You could take that extra comma and put it here, after “P-PCE”.

   Additionally a per-domain stitched LSP

And then make a copy of that comma and put it after “Additionally”.

   setup, whereas Section 3.3.1 describe the per-domain stitching.


— Section 1.2.1 —

   support the function of multi domain coordination via hierarchy,

Hyphenate “multi-domain”, or make it one word (“multidomain”).

   Virtual Network (VN) requirements before requesting for the VN to be

That sounds odd using “for” and “to be”.  It should be “…requesting that the VN be provisioned.”

   P-PCE and C-PCEs. When the CNC requests for VN provisioning, the MDSC

Remove “for”.

   decompose this request into multiple inter-domain LSP provisioning


— Section 1.2.2 —

   In case of a stateful P-PCE, the stateful P-PCE has to be aware of

You don’t have to say “stateful P-PCE” twice.  You can either remove the word “stateful” the second time, or (better, I think) simply remove the first clause and say, “A stateful P-PCE has to…”

   For example, a domain diverse path from another LSP.

That’s not a sentence; please make it one or merge it into an adjacent sentence (I can’t figure out the right way to do that from the text you have).

   The other LSP could be
   an associated LSP (such as protection) or an unrelated LSP whose

I don’t understand the use of “protection” here; can you clarify?

— Section 1.2.3 —

   Similarly, a P-PCE could also request for delegation if it needs to

Remove “for”.

— Section 3 —

   All PCE types herein (i.e., EC-EP or EP-EC) are assumed to

It should be “and”, not “or”, and I suggest removing the “i.e.” and just saying “(EC-EP and EP-EC)”.

   A number of interactions are expected in the Hierarchical Stateful
   PCE architecture, these include:

This is called a “comma splice”.  Make it a period instead (and have “These” start a new sentence).  Alternatively, change “these include” to “including”.

   Note that this hierarchy is recursive and thus a Label Switching
   Router (LSR), as a PCC could delegate the control to a PCE, which may
   delegate to its parent, which may further delegate it to its parent
   (if it exist or needed). Similarly update operations could also be
   applied recursively.

This has a number of problems, so let me just fix it here:

   Note that this hierarchy is recursive, so a Label Switching Router
   (LSR), as a PCC, could delegate control to a PCE, which may in turn
   delegate to its parent, which may further delegate to its parent
   (if it exists). Similarly, update operations can also be applied

   an Open message indicates the support for stateful H-PCE operations
   as described in this document.

Remove “as”.

   procedures to minimise computational loops.

“minimize” — if you use British spelling you need to be consistent about it, and you’re otherwise using American spelling throughout the document.

— Section 3.1 —

   It then sends computation requests to the C-
   PCEs responsible for each of the domains on the candidate domain

I think it’s better not to allow “C-PCE” (and other such terms) to be hyphen-split across lines.

   A local policy at
   C-PCE may dictate which LSPs to be reported to the P-PCE.

“to be” is not right here.  I’d make it “are” (or “should be”).

   State synchronization mechanism as described in [RFC8231] and
   [RFC8232] are applicable


   We use the sample hierarchical domain topology example from [RFC6805]

I’d remove “sample”, because “example” covers it.  Or remove “example”… either way.

   following additional steps are added for stateful PCE to be executed
   at the end:

Comma after “PCE”, please.

— Section 3.2 —

   As per [RFC8051], Delegation is an operation to grant a PCE,
   temporary rights

Remove the comma after “PCE”.

   The C-PCE may further choose to delegate to P-PCE based
   on a local policy.

To any P-PCE?  Or “to its P-PCE”?

   The PCRpt message with "D" (delegate) flag is
   sent from C-PCE to P-PCE.

‘with the “D” (delegate) flag’ (add ‘the’)

   For LSP delegated to the P-PCE via the child PCE,
   the P-PCE can use the same PCUpd message to request change to the C-
   PCE (the Ingress domain PCE), the PCE further propagates the update
   request to the PCC.

Again, multiple issues:

   For an LSP delegated to a P-PCE via the child PCE, the P-PCE
   can use the same PCUpd message to request a change to the C-PCE
   (the Ingress domain PCE).  The PCE further propagates the update
   request to the PCC.

— Section 3.3 —

   In case of inter-domain LSP in
   Hierarchical PCE architecture, the initiation operations can be
   carried out at the P-PCE.  In which case after P-PCE finishes the E2E
   path computation, it can send the PCInitiate message to the C-PCE

There are articles missing here.  I think it should be “an inter-domain LSP” and “after the P-PCE finishes”.

   The following steps are performed, for PCE initiated operations,

Remove the first comma and hyphenate “PCE-initiated”.

— Section 3.3.1 —

   operation is possible, where multiple intra-domain LSPs are initiated
   in each domain which are further stitched to form an E2E LSP.

Needs a comma after “each domain”.

   The following steps are performed, for the Per Domain stitched LSP

Remove the comma.

   Note that each per-domain LSP can be setup in parallel.

“set up”, two words.

— Section 5.1 —

   If a parent PCE receives report from an unauthorized child

“a report”

   All mechanism as described in
   [RFC8231] and [RFC8281] continue to apply.


— Section 5.2 —

   The PCEP YANG module
   [I-D.ietf-pce-pcep-yang] may be extended to include details for
   stateful H-PCE deployment and operation, exact attributes to be
   modeled is out of scope for this document.

Who will (or may) extend it, and when might that happen?
And change the comma to a period and capitalize “Exact” (another comma splice).

— Section 6.3 —

   Along with the confidentiality during path
   computation, the child PCE could also conceal the path information, a
   C-PCE may replace a path segment with a path-key [RFC5520],
   effectively hiding the content of a segment of a path.

Combination of problems:

   Along with maintaining confidentiality during path
   computation, the child PCE could also conceal the path information.
   A C-PCE may replace a path segment with a path-key [RFC5520],
   effectively hiding the content of a segment of a path.

Alvaro Retana No Objection

(Adam Roach) No Objection

Martin Vigoureux No Objection

Comment (2019-09-17 for -13)

thank you for this document. I only have nit type comments:

s/hierarchy of stateful PCEs play a crucial role/hierarchy of stateful PCEs plays a crucial role/

   The ability to compute shortest constrained TE LSPs
not sure what you mean by "shortest constrained"

s/the MDSC decompose/the MDSC decomposes/

   Different signaling methods for inter-domain RSVP-TE signaling 

s/the PCE further propagates/the C-PCE further propagates/

Éric Vyncke No Objection

(Magnus Westerlund) No Objection