Re: [secdir] secdir review of draft-ietf-mpls-rsvp-te-no-php-oob-mapping-08

["Adrian Farrel" <[e-mail]>]

Since the authors and document shepherd haven't responded, and since Stephen
mentions the review in Comment, I will take on responding...

> I have one minor question; in section 2.2 is this:
> 
>       An Ingress LSR sets the OOB mapping indication flag to signal the
>       Egress LSR that binding of RSVP-TE LSP to an application and
>       payload identification is being signaled out-of-band. This flag
>       MUST NOT be modified by any other LSRs in the network. LSRs other
>       than the Egress LSRs SHOULD ignore this flag.
> 
> On that last "SHOULD": what does it mean for any other LSR *not* to
> ignore the flag?  That is, what can they do?  How can they not ignore
> it, since there's no defined behaviour for them to do with it?

There is a difference between not being told to do something, and being told to
not do something.

It would be extreme, IMHO, to say that a transit LSR MUST ignore the flag.
The fact that there is no behavior required of the transit LSR and nothing that
pertains to the LSP that can be thought of for the LSR to do, is not reason to
forbid the LSR from looking at the flag, saying "Ooooh, that's interesting", and
sending a message to its third cousin in Baltimore to gossip about the fact.
In the same way that a router "SHOULD" ignore the source IP address on a packet
when it routes it, there is no reason to prohibit examination of the field.

Cheers,
Adrian