Re: [TLS] [pkix] Cert Enumeration and Key Assurance With DNSSEC

[Phillip Hallam-Baker <[e-mail]>]

On Fri, Oct 1, 2010 at 6:05 PM, Matt McCutchen <[e-mail]>wrote:

> On Fri, 2010-10-01 at 11:29 -0400, Phillip Hallam-Baker wrote:
> > In particular I am very concerned about the particular approach being
> > taken to security policy. What the proposers are attempting to do is
> > to create a mechanism that allows a site that only uses one particular
> > high assurance CA to 'protect' themselves against SSL certificates
> > being issued by low assurance CAs.
> >
> > As such, this is an objective I approve of and is one that I would
> > like to see supported in a generalized security policy. It should be
> > possible for a site to make security policy statements of the form
> > 'all valid PKIX certs for example.com have cert X in the validation
> > path'.
> >
> > What I object to is the approach being taken which is to use DNSSEC to
> > replace PKIX certificate validation entirely.
>
> Realize that I, and I would guess many other site admins, want precisely
> that.  PKIX is complicated, whereas once I have a DNSSEC signed zone,
> placing my TLS server's certificate in the zone and knowing that clients
> will accept that certificate and no other could hardly be simpler.  And
> why shouldn't I be allowed to do it?  I have complete authority over my
> zone (even for the most part in the present public CA system).  Nobody
> gave PKIX a monopoly on the determination of certificate acceptability.
>
> We could support a more general scheme in which positive assurance is
> separate from restrictions, but don't be surprised when a significant
> fraction of sites use it to effectively "replace PKIX certificate
> validation".
>
>
The problem with that approach is that the attacker now has two
infrastructures that they can attack rather than just one.

You are increasing your attack surface, not reducing and simplifying it as
you might imagine.


> Worse still, the proponents refuse to allow any method of shutting
> > this system off. So if I have a site where I want to use DNSSEC
> > validated certificates on the mail server, deployment is going to
> > impact my Web server.
>
> Yes, there should be a way to make the exclusivity optional, but there
> may be better ways to solve the problem you cited, such as placing the
> DNSSEC certificate at the SRVName for the mail server.
>


Regardless, I believe that the PKIX and TLS groups should be aware that this
is the change that has been implemented in code and is being proposed before
the WG is chartered.

There are much better ways to express the trust restriction semantics. I do
not see why the ability to express statements concerning your trust roots
needs to require deployment of a completely different mechanism and trust
path for the end entity keys.


For example, the ESRV mechanism is extensible. So it is possible for a site
that has their own PKI and root of trust to specify that there has to be a
trust path that includes that root of trust. Or you could publish a
fingerprint of a cross cert that has to be involved or you could even state
that there must be a cert record with the right fingerprint.


-- 
Website: http://hallambaker.com/