The SNMP architecture in RFC3411 has 'Abstract Service Interfaces'
between 'Subsystems'. What we are talking about here is a possible
'extension point' within a specific Model implementing a specific
Subsystem. I think these are different things and it remains unclear
whether it is worth to define 'extension points' of Models (USM is a
Model implementing the Security Subsystem Abstract Service Interface).

At least, we should not confuse 'Abstract Service Interfaces',
'Subsystems', 'Models' and 'extension points' (which is a new concept
since so far Models do not have such plugin extension points).

/js

On Thu, Aug 28, 2014 at 01:33:54PM -0400, Sam Hartman wrote:
> Hi.
> 
> My concern in describing things as diffs to the md5 algorithm in 3414 s
> that there's not a well defined abstraction there at that layer.
> 
> The SNMP community has been very careful to define rules of procedure
> and very well-defined interfaces afor extension and for variability in
> SNMP.  When we started work on the transport security model we
> considered providing differences to the existing SNMP processing.
> It would have been a lot less text.
> However, we  were not using proper SNMP extension points in doing so.
> Extending SNMP only at its proper extension points helps in reasoning
> about SNMP correctness, security, etc.
> For the same reason abstraction helps you when coding, it helps you when
> working on specifications for large complex systems.
> 
> Authentication algorithms in USM are a valid extension point today.
> However, as I read 3414, HMAC-based algorithms that work approximately
> like hmac-md5 are *not* a well-defined extension point.
> 
> I think it is important to create a new extension point that plugs into
> the existing extension point of USM authentication algorithm and
> provides hmac-based authentication with an extensible hash function.
> 
> draft-hartman tries to do that today. draft-hmac does not.
> 
> I think it is  important  for SNMP forward evolution that we do that and
> only extend at maintainable extension points.
> For that reason, regardless of which draft the working group chooses, I
> believe it is important to first define a new extension point for
> generic hmac-based authentication algorithms that plug into the USM
> authentication algorithm extension point in 3414, and then to plug sha-2
> into that newly created extension point.
> 
> --Sam
> 
> _______________________________________________
> OPSAWG mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/opsawg

-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: +49 421 200 3587         Campus Ring 1, 28759 Bremen, Germany
Fax:   +49 421 200 3103         <http://www.jacobs-university.de/>

_______________________________________________
OPSAWG mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/opsawg

Reply via email to