Draft:  draft-ietf-pwe3-cw-05.txt 
Reviewer: john.loughney@nokia.com
Review Date: Wednesday 8/17/2005 2:38 AM
Telechat Date: Thursday 8/18/2005

Summary: I think that the document is almost there, under technical 
comments, I highlight 3 issues that I think would enhance the current document.  In summary, they are:

1) Expand acronyms in the abstract.
2) Add some additional text around the SHOULDs in sections 2 & 3.
3) Add some text about potential security threats with this mechanism.

Technical Comments:
-------------------

1) I'm not an MPLS expert, so reading the abstract (two sentences) with
a large number of acronyms presents a bit of a barrier for understanding
exactly what is going on here:

   ... The 
   design of these fields is chosen so that an MPLS LSR performing MPLS 
   payload inspection will not confuse a PWE3 payload with an IP 
   payload. 

Suggestion: expand most of the acronyms.

2) Section 2 says:

   A PW carried over an MPLS PSN that uses the contents of the MPLS 
   payload to select the ECMP path SHOULD employ the PW MPLS Control 
   Word described in Section 3 for data, or the PW Associated Channel 
   Header described in Section 4 for channel associated traffic.

Why SHOULD it employ the PW MPLS control word or PW associated channel
header?  What if an implementation doesn't?  

3) In section 3, you have:

   When a PWMCW is used, it MUST adhere to the Generic MPLS Control 
   Word format as illustrated in Figure 1 above. It SHOULD also follow 
   the following format:  

What if it doesn't follow the format? Why would an implementation want
to follow the format?

4) Security Considerations section doesn't really address how a PW
Associated Channel could be abused, which I think makes it harder for
someone reading this document to be able to deterime what security
considerations they should take when deploying this.  I think some text
explaining what the threats are would be useful