Re: [Teas-ns-dt] New Version Notification for draft-nsdt-teas-transport-slice-definition-00.txt

[Eric Gray <[e-mail]>]

Draft authors,

Some comments on the -00 version of this ID are provided below.

As a couple of general observations:

  1.  This is a specification – hence precision is very important.
  2.  If you would like to use “transport slice” to mean “transport network slice” – you need to explicitly say that.

Major comments:

Minor comments:

The “Abstract” is awkward; I suggest rewording along the following lines…
    “This document provides a definition of transport network slicing in IETF, and describes
      considerations for realization of transport network slices.”

Justification: Definitions are typically not “described” (a description lists characteristics of a thing).  We should not refer to the design team’s role in writing this draft in the Abstract (note – the last sentence of the Abstract is deliberately omitted in the text I suggest for replacing the Abstract).  While we don’t usually include who is doing the work associated with a draft in the Abstract, the point to this effort is to define “transport network slices” (or “slicing”), not “transport slices.”  Also, I have already said that we should not conflate “implementation” with “realization.”

The first paragraph in the Introduction attempts to “define” network slicing.  In doing so, it is missing a lot.  I suggest replacing the entire paragraph (sans bullets) with:

    “Network slicing is a phrase that can be used to generically describe virtualization
      of network resources.  This phrase, therefore, is considered very useful by many
      people (and for many applications) because of the need to describe generically a
      diversity of services and related resource requirements that can then be applied to
      any of a number of proposed, implemented or deployed technology and associated
      devices.

    “Some key applications which might benefit from the use of network slicing include:”

<The bullet list provided is fine, with “NIT” exception listed below, but could include other applications.>

Justification:  The attraction for many in the use of the phrase “network slicing” is that it avoids including technology specifics that are pre-associated with existing technology approaches used for virtual networking, and associated terms/phrases/acronyms.  In a single phrase (“network slicing”) we include every technology that might be used to provide virtual networks (unlike VPN, VLAN, any of the many forms of traffic multiplexing, etc.).

Because “network slicing” can be used to generically describe any form of network virtualization, it naturally includes any form of traffic “separation,” “priority,” “differentiated services,” specialized queuing behaviors, etc. – that might be included in any technology that supports virtualization of networks.

Replace the last sentence in the Introduction (as proposed for the Abstract) with the following:
    “This document provides a definition of transport network slicing in IETF, and describes
      considerations for realization of transport network slices.”

In addition to the many NITs with the first sentence (mentioned below) of the first paragraph of section 2, it is a subjective statement that the network shown in figure 1 is “typical.”  It is a use case example that we want to include, but there may be many actual deployments that are simpler.

Figure 1 makes no reference to “operator-Y.”

In addition, it is very difficult to parse the second sentence in the first paragraph of section 2.  A network necessarily includes multiple nodes (not one or more).  The opening phrase ends with a preposition (“of”).  It is not at all clear that you are trying to include network functions, PNFs (?) and VNFs (?) as examples of “nodes” – if that is what you are trying to do.  I assume that PNF = physical network function, and VNF = virtual network function – but the sentence is too busy to just stick that in here.  And the phrase “which have various capabilities and technologies such as” does not seem to have enough value to be worth complicating the sentence.

The node examples are a mixed bag and it is not clear why this set provides a reasonable set of examples for the technologies already listed.

Also, you add in “aka” in several places, apparently with the meaning “includes.”  That is not what AKA means (it is the acronym for “also known as” – and you use this in cases where equivalency is not the case); as far as I am aware, there is no English word “aka” and – if there is – I don’t think it means “includes.”

I would suggest replacing the first two paragraphs (including the bullets in between) with the following:
   “To illustrate the IETF definitions of both E2E network slice and transport network slice,
     consider the network shown in Figure 1, where a network operator has an E2E network
     slice that traverses multiple technology-specific networks.

    “Each of these networks might use any of a number of technologies, including IP, MPLS,
      Fiber-Optics (e.g. – WDM, DWDM), Passive Optical Networking (PON), Microwave, etc.

    “Each of these networks also includes multiple (physical or virtual) nodes and might provide
      network functions beyond simply carrying of technology-specific protocol data units.

    “The types of nodes used in any of these networks may include:

  *   Packet Switches (e.g. – Routers, Bridges)
  *   Application servers
  *   Firewalls
  *   Radio Access Network (RAN) components
  *   Microwave transceivers
  *   Optical repeaters
  *   etc.

    “Each node in the network might support multiple technologies among those listed above.”

    “As an example combination of networks:”

<Again, the bullet list provided is fine, with “NIT” exception listed below>

In the second sentence after Figure 1, the adjectives “separate” and “independent” are ambiguous and add no value.  They should be omitted.

In the second paragraph after Figure 1, the second and third sentences are not consistent with each other.  An E2E slice is associated [with] “a customer” (singular), yet there is only a single E2E context between the two users EU-x and EU-y.  What are you trying to say?

Also, in the same paragraph, you still are abusing the acronym “SLA.”  Reword the afflicted sentence as:
    “… with a resource requirement for bandwidth greater than 10Mbps.”

Note that this appears to make the probably unreasonable assumption that the connection in this example from each CCTV camera to a central collection/observation facility is something that the operator can provide as a fully-connected hub and spoke network.  But the example is extremely ambiguous, because it is as likely to be interpreted as a fully-connected mesh network.

Whatever the use case really is, it is likely that the actual requirement is for some minimum bandwidth to be available for carrying video signal from a CCTV to some recording and/or observation facility – however that might be provided by the operator.


NITs:
In Introduction, first bullet “slicig” --> “slicing”
Second paragraph, “parts such as” --> “parts, such as”
Same paragraph, “established on” --> “established for”
Last paragraph in introduction (if not replaced as suggested above),

  *   “transport slice” --> “a transport network slice”
  *   “from IETF aspect” --> “from an IETF perspective,” (include the comma)
  *   “on their realization” --> “for transport network slice realization” (“their” is ambiguous)


In section 2, “illustrate” (or “depict”) would be better than “demonstrate” and it looks like there are two things being defined, hence “definition” should probably be plural – also you need to use either “an” or “the” in front of “IETF.”



In the bullet list toward the bottom of page 3, why are the numbered networks not listed in order?



In Figure 1, “Legends” --> “Legend” and “End-2-end” --> “End-to-end” – also, do we need to add EU-y?



In the first sentence after Figure 1, the Figure does not identify the operator; also in the text around the figure, it is not clear that there is any implication that there is more than one operator – hence a label is not needed.



In the second sentence after the figure, “its” is ambiguous – replace with “the operator’s.”  And “for specific service” --> “for a specific service.”



First sentence of the second paragraph after Figure 1, omit “As shown” at the beginning of the sentence, “logical” --> “logically” and “i ndependent” --> “independent.”



Second sentence, same paragraph – “associated to” --> “associated with”



Last paragraph on page 4, “associate to” --> “associated with”



I have other things I need to do, in spite of the importance of this work, so I will have to stop at the end of page 4.  I took a quick look and it seems to me that I had made a number of other comments using the Word version – which, if addressed, will likely clear up a number of the remaining issues with this document.