Ballot for draft-ietf-teas-rfc8776-update
Discuss
Yes
No Objection
Summary: Has 2 DISCUSSes. Has enough positions to pass once DISCUSS positions are resolved.
# Éric Vyncke INT AD comments for draft-ietf-teas-rfc8776-update-20 CC @evyncke Thank you for the work put into this document. Please find below one blocking DISCUSS points, some non-blocking COMMENT points/nits (replies would be appreciated even if only for my own education). Special thanks to Oscar Gonzalez de Dios for the shepherd's write-up including the WG consensus and the justification of the intended status. I hope that this review helps to improve the document, Regards, -éric Note: this ballot comments follow the Markdown syntax of https://github.com/mnot/ietf-comments/tree/main, i.e., they can be processed by a tool to create github issues. ## DISCUSS (blocking) As noted in https://datatracker.ietf.org/doc/statement-iesg-handling-ballot-positions-20220121/, a DISCUSS ballot is a request to have a discussion on the points below; I really think that the document would be improved with a change here, but can be convinced otherwise. ### Title While I understand that the TEAS WG is limited to MPLS technology, the title of this *IETF* draft is misleading as it it not `Common YANG Data Types for Traffic Engineering` but more about `Common YANG Data Types for *MPLS-based* Traffic Engineering`. This also applies to the module names as they should include "mpls" in the name as they are *specific* to MPLS and in no way `ietf-te-types`, or did I miss something? Even if RFC 8776 had the some issue, this is not a reason for repeating the confusion. More broadly, I find extremely sad that there is IETF-wide effort to have a real common data model across all TE control planes :-( ### Section 3.1 The leading text contains `TE types that are independent and agnostic of any specific technology or control-plane instance` but its subsection has terms that are very specific to MPLS-based control planes, e.g., `lsp-encoding-types`. ### Section 3.1.1 Missing text about when the SHOULD can be bypassed or what are the consequences of bypassing it in: `The derived identities SHOULD describe the unit and maximum value of the path metric types they define` per https://datatracker.ietf.org/doc/statement-iesg-statement-on-clarifying-the-use-of-bcp-14-key-words/ ### Section 3.1.2 What is `16-octet in full, mixed, shortened, or shortened-mixed IPv6 address notation` ? Why not simply requesting to use RFC 5952 ? Same applies to `te-node-id` in section ### Section 4 It is unclear what is "hex float" in `where each of these numbers can be non-negative decimal, hex integer, or *hex float*` in the description of`te-bandwidth`. There is a hint to `'bandwidth-ieee-float32'` but it should not be a hint.
## COMMENTS (non-blocking) ### Section 2 It would have been useful to also provide references to the terms beyond acronyms expansions. ### Section 3.1.1.1 Please expand or add a reference to "RWA" in `No RWA constraints met`
"Abstract", paragraph 0 > This document defines a collection of common data types, identities, > and groupings in YANG data modeling language. These derived common > data types, identities and groupings are intended to be imported by > other modules, e.g., those which model the Traffic Engineering (TE) > configuration and state capabilities. I have to agree with Eric and Joel Halpern (BTW, thanks, Joel, for your GENART review) about the fact that most of these definitions are, if anything, technology-specific, and the draft needs to reflect that. It is not enough to say that the definitions in this draft "can be used by any module" and, by implication, with other technologies, unless those technologies are also covered comprehensively by this module. Certainly, the assertion that the module is technology agnostic needs to be scrutinized. If the authors agree that the definitions are MPLS specific, I would not suggest renaming the module, as that is a much bigger change and will not play well with obsoleting RFC 8776. I would rather just update the description in the draft and in the YANG module to reflect that fact, and hope one day the module will be truly technology agnostic. Section 4, paragraph 21 > This revision obsoletes the following identities: > - of-minimize-agg-bandwidth-consumption; > - of-minimize-load-most-loaded-link; > - of-minimize-cost-path-set; > - lsp-protection-reroute-extra; > - lsp-protection-reroute. draft-ietf-netmod-rfc8407bis in Section 4.7 says that: The status SHOULD NOT be changed from "current" directly to "obsolete". An object SHOULD be available for at least one year after the publication date with a "deprecated" status before it is changed to "obsolete". Yet, this document has clearly done exactly that. Was it not possible to deprecate the above identities, introduce the new identities, and a year later, obsolete the identities that were deprecated? Section 6, paragraph 5 > This document requests IANA to register the following YANG modules in > the "YANG Module Names" registry [RFC6020][RFC9890] within the "YANG > Parameters" registry group. > > name: ietf-te-types > Maintained by IANA? N > namespace: urn:ietf:params:xml:ns:yang:ietf-te-types > prefix: te-types > reference: RFC XXXX > > name: ietf-te-packet-types > Maintained by IANA? N > namespace: urn:ietf:params:xml:ns:yang:ietf-te-packet-types > prefix: te-packet-types > reference: RFC XXXX Why is this registration request being made? Should the request not be for an update to refer to this document? Section 7, paragraph 5 > The YANG modules define a set of identities, types, and groupings. > These nodes are intended to be reused by other YANG modules. The > modules by themselves do not expose any data nodes that are writable, > data nodes that contain read-only state, or RPCs. As such, there are > no additional security issues related to the YANG module that need to > be considered. > > Modules that use the groupings that are defined in this document > should identify the corresponding security considerations. For > example using 'explicit-route-hop', 'record-route-state' or 'te- > topology-identifier' (which includes the 'client-id') groupings may > expose sensitive topology information. If there are no security issues that need to be discussed here, why is the rest of the text needed? Just these paragraphs should suffice, no?
Section 1.2, paragraph 1 > +=================+======================+========================+ > | Prefix | YANG module | Reference | > +=================+======================+========================+ > | yang | ietf-yang-types | Section 3 of [RFC6991] | > +-----------------+----------------------+------------------------+ > | inet | ietf-inet-types | Section 4 of [RFC6991] | > +-----------------+----------------------+------------------------+ > | rt-types | ietf-routing-types | [RFC8294] | > +-----------------+----------------------+------------------------+ > | te-types | ietf-te-types | RFCXXXX | > +-----------------+----------------------+------------------------+ > | te-packet-types | ietf-te-packet-types | RFCXXXX | > +-----------------+----------------------+------------------------+ Please update all references to RFC6991 to point to RFC 9911. Section 4, paragraph 14 > All revisions of IETF and IANA published modules can be found > at the YANG Parameters registry group > (https://www.iana.org/assignments/yang-parameters). Is this draft publishing an IANA module? If not, please remove reference to IANA modules. Section 4, paragraph 34 > identity lsp-provisioning-error-reason { > description > "Base identity for LSP provisioning errors."; > } I do not see this base identity being used in this module or in this document. If this is intended to be used somewhere else or at a later time, it would be better to define it there or at that time. Section 4, paragraph 358 > leaf one-way-delay { > type uint32 { > range "0..16777215"; > } > units "microseconds"; > description > "One-way delay or latency."; > } Is there a reference that can be cited for this? Or even an explanation for how it can be measured. Per MEF 10.3, "One-way delay is difficult to measure and therefore one-way delay may be approximated from two-way measurements." Section 4, paragraph 357 > leaf one-way-delay-normality { > type te-types:performance-metrics-normality; > description > "One-way delay normality."; > } This description and several like these are essentially a repeat of the name of the node without an explanation for what it is. Either an explanation or a reference is the minimum one should expect from these modules. Section 4, paragraph 358 > leaf two-way-delay { > type uint32 { > range "0..16777215"; > } > units "microseconds"; > description > "Two-way delay or latency."; > } Same question here for reference. Section 5, paragraph 0 > The "ietf-te-packet-types" module imports from the "ietf-te-types" > module defined in Section 4 of this document. How about RFC 6991 or more precisely RFC 9911? Section 5, paragraph 1 > // RFC Editor: Please replace XXXX above with the RFC number > // assigned to this document. > // Please remove this note. Can all these RFC Editor directives be consolidated in one place instead of scattering them across the document? Section 5, paragraph 54 > grouping bandwidth-profile-parameters { > description > "Common parameters to define bandwidth profiles in packet > networks."; > leaf cir { > type uint64; > units "bits per second"; > description > "Committed Information Rate (CIR)."; > } > leaf cbs { > type uint64; > units "bytes"; > description > "Committed Burst Size (CBS)."; > } > leaf eir { > type uint64; > units "bits per second"; > description > "Excess Information Rate (EIR)."; > } > leaf ebs { > type uint64; > units "bytes"; > description > "Excess Burst Size (EBS)."; > } > leaf pir { > type uint64; > units "bits per second"; > description > "Peak Information Rate (PIR)."; > } > leaf pbs { > type uint64; > units "bytes"; > description > "Peak Burst Size (PBS)."; > } > } Please cite the reference of MEF 10.3 for these parameters as suggested by Sergio Belotti. BTW, thanks, Sergio, for your OPSDIR review. Found terminology that should be reviewed for inclusivity; see https://www.rfc-editor.org/part2/#inclusive_language for background and more guidance: * Term "he"; alternatives might be "they", "them", "their" ------------------------------------------------------------------------------- NIT ------------------------------------------------------------------------------- All comments below are about very minor potential issues that you may choose to address in some way - or ignore - as you see fit. Some were flagged by automated tools (via https://github.com/larseggert/ietf-reviewtool), so there will likely be some false positives. There is no need to let me know what you did with these suggestions. Reference [RFC6991] to RFC6991, which was obsoleted by RFC9911 (this may be on purpose). "I", paragraph 53 > with the alternate route being pre-computed and stored for use when the fai > ^^^^^^^^^^^^ Do not mix variants of the same word ("pre-compute" and "precompute") within a single text. "I", paragraph 196 > es a type representing the access type of a TE link."; reference "RFC 3630: T > ^^^^^^^^^ If "type" is a classification term, "a" is not necessary. Use "type of". (The phrases "kind of" and "sort of" are informal if they mean "to some extent".). "I", paragraph 210 > ition, and the user traffic is being transported (or selected) on the recove > ^^^^^^^^^^^^^^^^^^^^ You have used the passive voice repeatedly in nearby sentences. To make your writing clearer and easier to read, consider using active voice. "I", paragraph 333 > address or dotted quad address or as an URI. The type data node disambiguate > ^^ Use "a" instead of "an" if the following word doesn't start with a vowel sound, e.g. "a sentence", "a university". "I", paragraph 397 > c-editor.org/rfc/rfc9522>. Appendix A. The Complete Schema Trees This appendi > ^^ It seems like there are too many consecutive spaces here. Section 8, paragraph 10 > ready defined in [RFC8776], have been updated in the 'explicit-route-hop': - > ^^^^^^^^^^^^ You have used the passive voice repeatedly in nearby sentences. To make your writing clearer and easier to read, consider using active voice. Section 8, paragraph 11 > n replaced by must statements that requires at least the presence of: o node > ^^^^^^^^ Possible subject-verb agreement error detected. Section 8, paragraph 12 > n replaced by must statements that requires at least the presence of: o node > ^^^^^^^^ Possible subject-verb agreement error detected. Section 8, paragraph 14 > constraints; The following new leaf have been added to the 'tunnel-constraint > ^^^^ The verb form "have" does not seem to match the subject "leaf". Section 8, paragraph 15 > jects-always; The container has been renamed as 'explicit-route-objects'. Thi > ^^^^^^^^^^^^ You have used the passive voice repeatedly in nearby sentences. To make your writing clearer and easier to read, consider using active voice.
Hi Italo, Aihua, Xufeng, Tarek, and Igor, Thank you for the effort put into this bis and specifically the meticulous work to call out all changes and their rationale. The reasoning for not following some of the RFC8407 guidance are well-explained. Likewise, explanations about why an IANA-maintained module is not used for PCE errors are reasonable. Thanks also to Sergio for the OPSDIR review. I trust the authors will follow-up. I reviewed this document several times in the past, so I only have nits that I will submit as PR for the authors convenience. The NMDA mention inside the modules (and also in the document) is useless. The new guidance is to include a justification only when there is major deviation. Cheers, Med
Thanks to Tero Kivinen for their secdir review. I agree with his assessment of the HUGE number of referenced RFCs. (16 references for 1 definition seems excessive)
# Internet AD comments for draft-ietf-teas-rfc8776-update-19 CC @ekline * comment syntax: - https://github.com/mnot/ietf-comments/blob/main/format.md * "Handling Ballot Positions": - https://ietf.org/about/groups/iesg/statements/handling-ballot-positions/ ## Comments * For anyone else looking to review a diff, this URL was a helpful starting place: https://author-tools.ietf.org/iddiff?url1=rfc8776&url2=draft-ietf-teas-rfc8776-update-19&difftype=--html
I support Mahesh's DISCUSS position.
I agree with the SecDir readability comments noted by Tero.
Thank you to Joel Halpern for the GENART review . I support the DISCUSS point of Mahesh Jethanandani, also identified by the GENART review, to clarify the relationship of this model to MPLS.