Re: [pcp] Review of draft-ietf-pcp-proxy-04

[<[e-mail]>]

Hi Tiru,

Thanks for the comments.

Please see inline.
Cheers,
Med

De : Tirumaleswar Reddy (tireddy) [mailto:[e-mail]]
Envoyé : lundi 20 janvier 2014 13:28
À : BOUCADAIR Mohamed IMT/OLN
Cc : Dan Wing (dwing); Reinaldo Penno (repenno); [e-mail]
Objet : Review of draft-ietf-pcp-proxy-04

Hi Med,

Comments:

a)
If PCP authentication is used, Is there a need to still use THIRD_PARTY option for the use case "No NAT is Co-located with the PCP Proxy"  in section 6 of the draft ?
[Med] THIRD_PARTY option is needed even if no NAT is co-located with the PCP Proxy (e.g., DS-Lite context). For other contexts than DS-Lite, if the source IP address is not altered by the PCP proxy, the proxy has to inspect all incoming packets to find out pcp ones and therefore enforces the appropriate policies. Altering the source IP address by the pcp proxy is a solution to redirect all the incoming  pcp traffic explicitly to the PCP proxy. In addition, as indicated below, proxy implications on authentication are out of scope of this document. The proxy document adopts the same rationale as the base pcp specification.

b)
Are you planning to update the draft  with details about the usage of PCP proxy with PCP auth.
[Med] NO. The plan is to follow what have been presented during the Berlin meeting (http://www.ietf.org/proceedings/87/slides/slides-87-pcp-2.pdf): "Proxy implication on authentication are out of scope and should be specified in pcp-authentication document"

In addition to the existing use cases this integration will be useful for http://tools.ietf.org/html/draft-wing-pcp-flowdata-00 where the PCP proxy in the Home network receives the flow characteristics signaled by the PCP client and propagates the flow characteristics to the PCP Server in the Access Network.

Cheers,
-Tiru.