[Cats] Progress of draft-yl-cats-data-model

[Adrian Farrel <[e-mail]>]

Hi,

 

I want to encourage more reviews of this document. Although this

document is at -07, it seems to me that there has been very little

discussion of the document on the list.

 

I have asked for a YANG Doctor preliminary review, and my own review

follows. But I would ask others on the list to make some time for this

draft.

 

Best,

Adrian

 

===

 

1.

 

I think you should repeat the paragraph about NMDA from section 3.2.

 

---

 

I'm not very clear why this document introduces the CATS-SBI and C-SMA

API. I see what they are for, and I read the text at the top of section

3.3. But nothing later in the document mentions these two interfaces or

shows how the YANG models are used on those interfaces.

 

I think you either drop all mention of the interfaces in the document,

or you should add a couple of sections on "Applicability of These YANG

Models to the <interface>"

 

---

 

3.3.4.

 

Isn't there a risk of a flap situation where one of the limits is 

reached and unreached in a quick repetition. This would cause a flood of

notifications. A way to protect against this is to set a rate-limit 

object.

 

---

 

4.

 

I don't think this document is the right reference for cs-id and csci-id

You should reference the document that defines those concepts.

 

---

 

4.

 

The description for "leaf matched-bytes" seems wrong. Perhaps...

 

               description

                 "The number of bytes in packets matching the specified

                  service.";

 

---

 

4.

 

It seems that the values of 'priority' and 'affinity' have opposite

interpretations: 0 is high priority, 0 is low affinity. No reason to

change this, but just wondered whether consistency would be helpful.

 

---

 

4.

 

In 'ccntainer metrics' you need to explain what YYYY is.

 

---

 

It would be nice to add an Operational Considerations section per

draft-ietf-opsawg-rfc5706.