Re: [mpls] LDP Security

["t.petch" <[e-mail]>]

----- Original Message -----
From: "Susan Hares" <[e-mail]>
Sent: Saturday, November 11, 2017 3:58 AM

> Kathleen and Eric:
>
> I agree that getting linux implementations with TCP-AO is one avenue
to drive things into product.   Any ideas on getting more deployment for
routing security - is really important.   Pushing both small and large
gains (TCP-AO, registration) will help us continually increase the
security.
>

After a lifetime of trying to forestall disasters, and failing, only to
have clients say afterwards 'I wish we had listened to you', my
conclusion is that the disaster has to happen first; only then will
users be motivated enough to incur the cost, the hassle of changing
their ways.

I see this with systems and networks, and I hear of just such a case
which led to the present level of security being added to BGP.

What we should - must - do is ensure that when disaster strikes, and
users are motivated, then the specifications are ready; two stalled
I-Ds, as was mentioned up-thread, is not ready.

Tom Petch

> Sue
>
> -----Original Message-----
> From: Kathleen Moriarty [mailto:[e-mail]]
> Sent: Thursday, November 9, 2017 3:09 PM
>

> On Thu, Nov 9, 2017 at 2:09 PM, Eric Rescorla <[e-mail]> wrote:
> > Yeah, I agree. I don't really have any good ideas how to get people
to
> > do AO. Based on comments I've heard, providers don't see a lot of
> > value, rightly or wrongly....
>
> I know Alia has been talking about a possible Linux implementation to
drive the way.  I agree with EKR on other points and have been following
along.
>
> Best,
> Kathleen
> >
> > -Ekr
> >
> > On Wed, Nov 8, 2017 at 9:15 PM, Susan Hares <[e-mail]> wrote:
> >>
> >> Eric:
> >>
> >> BGP and LDP would be more secure if TCP-AO deployed with all BGP
and
> >> LDP – but there are issues with customer pick-up and deployment of
> >> these protocols on many networks.  I wished we had TCP-AO when BGP
started, but we did not.
> >>
> >> Some of the least secure BGP is in data centers – where the DC
> >> providers say “It’s all under one administration”.  Another problem
is on private
> >> lines.    We should chat about the networks each of these protocols
are
> >> actually deployed on.   If you have any insight on a way to
encourage
> >> adoption, I’d love to hear it. Require TCP-AO does not really mean
> >> anything if providers and Data Centers do not adopt it.
> >>
> >> Going from SHA-1 to SHA-256 on a TCP-AO is simple upgrade compared
to
> >> getting people to TCP-AO.
> >>
> >> Sue
> >>
> >> From: mpls [mailto:[e-mail]] On Behalf Of Eric
Rescorla
> >> Sent: Wednesday, November 8, 2017 7:44 PM
> >> To: Uma Chunduri
> >>
> >> I don't understand what you're getting at here. Yes, if people have
> >> TCP-AO then presumably they have SHA-1.
> >>
> >> But now we're talking about requiring people to have TCP-AO in this
> >> case, so we should try to move them to SHA-256 at the time we
require AO.
> >>
> >> -Ekr
> >>
> >> On Wed, Nov 8, 2017 at 4:14 PM, Uma Chunduri
> >> <[e-mail]>
> >> wrote:
> >>
> >> From: Eric Rescorla [mailto:[e-mail]]
> >> Sent: Wednesday, November 08, 2017 3:53 PM
> >>
> >> To: Uma Chunduri <[e-mail]>
> >>
> >> On Wed, Nov 8, 2017 at 3:50 PM, Uma Chunduri
> >> <[e-mail]>
> >> wrote:
> >>
> >> In-line [Uma1]:
> >>
> >> Uma C.
> >>
> >>
> >> From: Eric Rescorla [mailto:[e-mail]]
> >> Sent: Wednesday, November 08, 2017 12:53 PM
> >>
> >> On Wed, Nov 8, 2017 at 11:57 AM, Uma Chunduri
> >> <[e-mail]>
> >> wrote:
> >>
> >> Hi Stewart,
> >>
> >> I would note https://tools.ietf.org/html/rfc6952 - where LDP
security
> >> is analyzed from all aspects.
> >>
> >> Eric,
> >>
> >> Quick comments below [Uma]:
> >>
> >> Uma C.
> >>
> >> From: mpls [mailto:[e-mail]] On Behalf Of Eric
Rescorla
> >> Sent: Wednesday, November 08, 2017 10:00 AM
> >>
> >> Hi Stewart
> >>
> >> Thanks for your note.
> >>
> >> My overall sense of the state of play is, I think much like yours.
> >>
> >> TCP-MD5 is inadequate in two major respects:
> >>
> >> - It uses weak algorithms
> >>
> >> - It has a bad negotiation/setuop story (manual key management)
> >>
> >> TCP-AO is intended to be a drop-in replacement for TCP-MD5 and so
> >> remedies the algorithm
> >>
> >> Issue
> >>
> >> [Uma]: Yes, if we go with RFC 5926 mandatory list..
> >>
> >> but not the key management issue [0]. We haven't made much progress
> >> on the key
> >>
> >> management story, and that seems to be a major impediment to
> >> deploying either of these
> >>
> >> technologies (which I am given to understand don't see a lot of
use).
> >>
> >> [Uma]: True.
> >>
> >>                But I would indicate some effort done few years back
> >> regarding key management for pair wise routing protocols (BGP, LDP,
> >> PCEP, MSDP ..).
> >>
> >>                One such proposal is by extending IKEv2 to negotiate
> >> TCP-AO MKTs (which can give rekey & algo. agility) -
> >> https://tools.ietf.org/html/draft-mahesh-karp-rkmp-05
> >>
> >>                This also requires some more work with TCP-AO; me &
> >> Joe put together
> >>
https://www.ietf.org/archive/id/draft-chunduri-karp-using-ikev2-with-
> >> tcp-ao-06.txt
> >>
> >>            Note the above didn’t progress in the concluded KARP WG
> >> (not fully sure the reasons on why).
> >>
> >> Yeah, I know that people tried to do this, but my impression was it
> >> kinda didn't progress much.
> >>
> >> We should probably talk in Singapore about that, but that's not
going
> >> to get better any time soon.
> >>
> >> In the interim, I think the text you have is OK, and "TBD" should
> >> read "SHA-256", with
> >>
> >> the fallback being SHA-256 -> SHA-1 -> MD5.
> >>
> >> [Uma]: While the list can be extended - I didn’t see SHA256 in the
> >> mandatory list in RFC 5926 for MAC.
> >>
> >> Generally we're trying to move away from SHA-1 towards SHA-256.
> >>
> >> [Uma1]: Couple of things:
> >>
> >> 1.       Nothing to be done (from spec pov of course): Use TCP-AO
(instead
> >> of current MD5) with the RFC 5926 mandated MACs/KDFs – so the ‘TBD’
> >> in Stewart suggesting below is already there.
> >>
> >> 2.       As #1 too is not good enough from your above note - do
SHA-256
> >> and live with it (no algorithm agility). Still a security benefit
in
> >> one way from existing stuff or even  #1.
> >>
> >> I'm not sure why you say "no algorithm agility". You'd be using AO,
> >> just with a different algorithm than SHA-1. AES-CMAC is still fine
as
> >> far as I know.
> >>
> >> [Uma2]: Sure, you have it, if you use AO;
> >>
> >>                  But then  I am not getting how we can mandate one
> >> MUST implement algorithm as suggested below TBD  would actually
work
> >> (especially
> >> - *if* #1 is already deployed somewhere?)
> >>
> >>                  Perhaps staying with #1 is the best bet or do
> >> negotiation through #3, with already mandated and additional stuff.
> >>
> >> -Ekr
> >> 3.       Do key management and “theoretically” get all we wanted….
> >>
> >> We have been here multiple times; because #1 itself is not *mostly*
> >> deployed (neither in BGP nor in LDP) if there is any appetite for
#2
> >> and #3 for practical deployments. But still it may be good to do #2
any ways.
> >>
> >> -Ekr
> >>
> >> -Ekr
> >>
> >> [0] Technically It has better support for rollover, but this is not
a
> >> huge improvement.
> >>
> >> [1] tcpcrypt is kind of orthogonal here as it's unauthenticated but
> >> opportunistic.  That said,
> >>
> >> it would provide defense against attackers who gain access to the
> >> link after connection
> >>
> >> setup and doesn't require configuration.
> >>
> >> On Wed, Nov 8, 2017 at 9:27 AM, Stewart Bryant
> >> <[e-mail]>  wrote:
> >>
> >> To the SEC and RTG ADs,
> >>
> >> I am sending the following message on behalf of the MPLS and the
PALS
> >> WG Chairs.
> >>
> >> There is a concern shared among the security community and the
> >> working groups that develop the LDP protocol that LDP is no longer
> >> adequately secured. LDP currently relies on MD5 for cryptographic
> >> security of its messages, but MD5 is a hash function that is no
> >> longer considered to meet current security requirements.
> >>
> >> In RFC5036 (published 2007) Section 5.1 (Spoofing) , List element
2.
> >> Session communication carried by TCP the following statements is
made:
> >>
> >> "LDP specifies use of the TCP MD5 Signature Option to provide for
the
> >> authenticity and integrity of session messages.
> >>
> >> "[RFC2385] asserts that MD5 authentication is now considered by
some
> >> to be too weak for this application.  It also points out that a
> >> similar TCP option with a stronger hashing algorithm (it cites
SHA-1
> >> as an example) could be deployed.  To our knowledge, no such TCP
> >> option has been defined and deployed.  However, we note that LDP
can
> >> use whatever TCP message digest techniques are available, and when
> >> one stronger than MD5 is specified and implemented, upgrading LDP
to use it would be relatively straightforward."
> >>
> >> We note that BGP has already been through this process, and
replaced
> >> MD5 with TCP-AO in RFC 7454. I would be logical to follow the same
> >> approach to secure LDP. However, as far as we are able to
ascertain,
> >> there is currently no recommended, mandatory to implement,
> >> cryptographic function specified. We are concerned that without
such
> >> a mandatory function, implementations will simply fall back to MD5
> >> and we will be no further forward
> >>
> >> We think that the best way forward is to publish a draft similar to
> >> RFC
> >> 7454 that contains the following requirement:
> >>
> >> "Implementations conforming to this RFC MUST implement TCP-AO to
> >> secure the TCP sessions carrying LDP in addition to the currently
> >> required TCP MD5 Signature Option. Furthermore, the TBD
cryptographic
> >> mechanism must be implemented and provided to TCP-AO to secure LDP
> >> messages. The TBD mechanism is the preferred option, and MD5 is
only
> >> to be used when TBD is unavailable."
> >>
> >> We are not an experts on this part of the stack, but it seems that
> >> TCP security negotiation is still work in progress. If we are
wrong,
> >> then we need to include a requirement that such negotiation is also
> >> required. In the absence of a negotiation protocol, however, we
need
> >> to leave this as a configuration process until such time as the
> >> negotiation protocol work is complete. On completion of a suitable
> >> negotiation protocol we need to issue a further update requiring
its use.
> >>
> >> Additionally we should note that no cryptographic mechanism has an
> >> indefinite lifetime, and that implementation should note the IETF
> >> anticipates updating the default cryptographic mechanism over time.
> >>
> >> The TBD default security function will need to be chosen such that
it
> >> can reasonably be implemented on a typical router route processor,
> >> and which will provide adequate security without significantly
> >> degrading the convergence time of an LSR. Without a function that
> >> does not significantly impact router convergence we simply close
one
> >> vulnerability and open another.
> >>
> >> As experts on the LDP protocol, but not on security mechanisms, we
> >> need to ask the security area for a review of our proposed
approach,
> >> and help correcting any misunderstanding of the security issues or
> >> our misunderstanding of the existing security mechanisms. We also
> >> need the recommendations of a suitable security function (TBD in
the above text).
> >>
> >> Best regards
> >>
> >> The MPLS WG Chairs
> >> The PALS WG Chairs