Control Protocol Extensions for the Setup of Time-Division Multiplexing (TDM) Pseudowires in MPLS Networks
draft-ietf-pwe3-tdm-control-protocol-extensi-07

Review by Christian Vogt:

This Internet draft specifies the establishment of TDM pseudowires
over MPLS.  It defines the usage of existing protocols and information
elements for this purpose, as well as required extensions.

The document is complete in my non-expert view, although it should be
revised for clarity.  While likely understandible for all pseudowire
specialists, readers not directly involved in this engineering area
may need more guidance.  This is editorial only.  A few suggestions:

- Introduction:  The introduction mixes the description of the
  document scope, items not in scope of the document, and a survey
  of related documents.  Consider restructuring.  Also, for
  non-experts in things pseudowires, adding a problem statement
  would be helpful.

- Section 2:  The key message of this section is that certain existing
  pseudowire FECs can be reused for TDM pseudowire establishment,
  with some restrictions in the parameters used.  The message is
  somewhat lost throughout the section.  Suggestion: State this
  clearly already at the beginning of the section.

- Section 3.1:  The unit of interface parameter length in the table is
  different than the unit used in the text.  This is confusing
  because the table does not state its unit.  Suggestion: State the
  unit in the table and use the same unit both in the table and in
  the text.

- Section 3.2:  The terms "SAToP" and "CESoPSN" are used to refer to
  groups of pseudowire types.  These terms haven't been defined
  before.  So to avoid confusion or misunderstanding, I suggest to
  explicitly name the pseudowire types in question at this point. (I
  do understand that "SAToP" refers to pseudowires 0x0011, 0x0012,
  0x0013, 0x0014, and that "CESoPSN" refers to pseudowires 0x0015
  and 0x0017.  But I think it should be clarified.)

- Section 3.3 specifies for which pseudowire types the CEP/TDP Bit
  Rate parameter may be omitted (item 1 in numbered list).  Suggest
  to name the affected pseudowire type codes, perhaps in
  parentheses, just to avoid misunderstandings.

- Section 8:  The security considerations do not consider
  vulnerabilities that non-perfect emulation of a particular link
  layer (TDM, in this case) could introduce.  Legacy applications
  may rely on TDM-specific properties that the emulated version over
  MPLS does not provide.  If there are no such vulnerabilities,
  which seems likely, then this should be state.

Section 3.6: bit diagram has type 0x0F, IANA Considerations text
suggests value 0x11?

Needs a normative reference to RFC 2119.