Re: [secdir] Secdir last call review of draft-ietf-mpls-sfc-encapsulation-02

["Andrew G. Malis" <[e-mail]>]

Paul,

Thanks for your review!

Adrian,

I don't think I agree with your statement that the MPLS forwarding plane
doesn't offer any security features. As I discuss in the last paragraph in
the security considerations in the draft, it's resistant to some classes of
attacks such as injecting unexpected packets unless the attacker was an
insider or had specific inside knowledge, such as valid SFF labels that
would be accepted by a receiving node. But then, it's difficult to defend
against any sort of attack on a router by an insider, starting with having
access to the physical box and the CLI.

While there are no specific NSH security mechanisms, I don't think this
draft is the place to have that discussion, as it's an MPLS WG draft. The
draft does say that this encapsulation is no more or less secure than
carrying the NSH in any other encapsulation. It's up to the SFC WG to
figure out NSH security elsewhere.

Cheers,
Andy



On Sun, Feb 17, 2019 at 7:08 PM Adrian Farrel <[e-mail]> wrote:

> Hi Paul,
>
> Should we make the observation that, since the MPLS forwarding plane does
> not offer any security features, if security is required it must be
> provided by the SFC layer using some mechanisms inherent in the NSH. We
> could/should probably also observe that, as yet, no NSH security mechanisms
> have been defined.
>
> Cheers,
> Adrian
>
> -----Original Message-----
> From: ietf <[e-mail]> On Behalf Of Paul Wouters
> Sent: 17 February 2019 23:58
> To: [e-mail]
> Cc: [e-mail]; [e-mail];
> [e-mail]
> Subject: Secdir last call review of draft-ietf-mpls-sfc-encapsulation-02
>
> Reviewer: Paul Wouters
> Review result: Ready
>
> I have reviewed this document as part of the security directorate's
> ongoing effort to review all IETF documents being processed by the
> IESG.  These comments were written primarily for the benefit of the
> security area directors.  Document editors and WG chairs should treat
> these comments just like any other last call comments.
>
> The summary of the review is Ready
>
> While I'm not familiar with the Service Function Chaining (SFC)
> architecture
> and the Network Service Header (NSH), the Security Considerations in this
> document seem to be correct in pointing out that:
>
>   This document simply
>    defines one additional transport encapsulation.  The NSH was
>    specially constructed to be agnostic to its transport encapsulation.
>    As as result, in general this additional encapsulation is no more or
>    less secure than carrying the NSH in any other encapsulation.
>
>
>