RE: 3280 Bis and Trust Anchors

["Santosh Chokhani" <SChokhani@xxxxxxxxxxxx>]

So, from security standpoint do you agree with the analysis that there
is some value to checking the expiration date on a trust anchor?

-----Original Message-----
From: Thierry Moreau [mailto:thierry.moreau@xxxxxxxxxxxxx] 
Sent: Wednesday, April 16, 2008 12:59 PM
To: Santosh Chokhani
Cc: pkix
Subject: Re: 3280 Bis and Trust Anchors



Santosh Chokhani wrote:
> For a long period of time, I used to think that checking validity
period 
> and revocation status of a trust anchor is useless.
> 
>  
> 
> While I still hold that view on the revocation status, my rationale
for 
> checking validity period has been incomplete.
> 
>  
> 
> It seems that there is a value in enforcing the validity period on a 
> trust anchor.
> 
>  
> 
> Trust anchors by their very nature are insecure objects in the sense 
> that they must be protected using means other than signature on them.

> To ascribe security to PKI, one has to assume that the means to
protect 
> the trust anchors in relying party trust store are secure and can not
be 
> altered.  Thus, enforcing validity period on them gives the
organization 
> another means to obsolete them.  This may be useful capability as we 
> transition from 1024 bit roots to 2048 bits and 1024 bit roots have 
> defined validity period.
> 
>  
> 
> I doubt that X.509 and 3280bis would want to change their requirement,

> but I hope that there would be less of an objection to discuss this in

> the Security Considerations section.
> 

In other words, a PKI expert changes mind about an issue related to 
global trust dissemination in 2008. It strikes me that experts in the 
field had, for so long, so little foresight in the area of trust anchor 
key (TAK) management.

In the above instance, my work on this provides a clean path for a 
solution to the concern of TAK rollover for the purpose of increasing 
key sizes. I do not feel the need to explain further since a) the PKI 
technology as so many rebuttal facets, and b) I'm just fed up with the 
IETF standards drafting process.

Anyway, the expired draft is

draft-moreau-pkix-takrem-01.txt
Trust Anchor Key Renewal Method Applied to X.509 Self-signed 
Certificates (TAKREM-X.509)

Abstract


	In the Internet PKI, trust anchor keys are distributed as
self-signed 
X.509 security certificates. This document specifies a trust anchor key 
renewal mechanism that leverages the confidence in the initial 
certificate distribution. A non-critical X.509 certificate extension 
holds a sequence of opaque octet strings. The trust anchor renewal 
operation occurs upon receipt of a message that hashes to one of those 
octet strings.

Regards,


-- 

- Thierry Moreau

CONNOTECH Experts-conseils inc.
9130 Place de Montgolfier
Montreal, Qc
Canada   H2M 2A1

Tel.: (514)385-5691
Fax:  (514)385-5900

web site: http://www.connotech.com
e-mail: thierry.moreau@xxxxxxxxxxxxx