Early Review of draft-ietf-lime-yang-connectionless-oam-05
I am the assigned YANG Doctor for the LIME YANG modules, this is the notes from my initial review of draft-ietf-lime-yang-connectionless-oam done with the authors on a call last week:
The module is intended to be implemented as a service model, per the module classification draft (https://tools.ietf.org/html/draft-ietf-netmod-yang-model-classification-05) with augments to topologies, I suggest it needs more description on the general concepts and how to use it.
I suggest to remove the 'OAM data hierarchy' section. It does not add any information.
The current version YANG module is all configuration except the 'oper' subtree. And there are no mandatory nodes (except list keys) which seems to allow for many, many combinations that likely make no sense. Please re-review and make sure this is an informed choice.
Suggest to rename tp-address-type-value to tp-address-technology-type.
Why does 'grouping tp-technology' only allow 'technology-null' or 'technology-string'/'leaf ipv4-icmp' and what is the meaning of the leaf ipv4-icmp content?
There are a couple of occasions of 'ordered-by user' that does not have a description of the meaning of the order. Please review.
Add an if-feature on path-discovery in *-methods draft.
Generally I'd suggest adding more information about the various types and which specifications you are picking them up from (e.g. RD, LSP ID, etc)
The specification defines a typedef that is a path in another module ("routing-instance-ref"). Suggest to reach out to authors of draft-ietf-rtgwg-ni-model-02 to see if they want the -ref type in that module for reuse.
Generally check whitespace and comments. Use the '-f yang' option to YANG to fix whitespace.
Downcase "IPv4-Multicast-Group-Address", and "IPv6-Multicast-Group-Address", "IP-Multicast-Group-Address"
Name of top container ('oper') not very helpful, suggest renaming to 'session-statistics'?
Spelling of grouping (and uses) cc-session-statsitics
Filename should include revision in CODE BEGINS
Don't reference drafts, pleas check RFC6087 bis and well-formed drafts for header stmts