Re: [mpls] New version of draft-boutros-mpls-tp-loopback-01

[Ben Niven-Jenkins <benjamin.niven-jenkins at bt.com>]

Colleagues,

On 23/03/2009 17:01, "Sami Boutros" <sboutros at cisco.com> wrote:
>> 4) Section 4.5 - Auth TLV. If the message is sent in-band, do we need a
>> auth-tlv ? Was there any requirement (in some requirement draft) to have
>> auth for loopback mode?
> 
> Good point will ask to add this requirement to the requirement doc for OAM
> intrusive functions.

Although I haven't read the draft (sorry Sami!), I can see need for
authentication for loopback.

One use case I am looking at where I need performance monitoring includes
placing T-PEs on unsecured customer premises. Although not foolproof use of
authentication may be required to lessen exposure of my network to attack
vectors through the performance monitoring mechanism.

Ben