Re: [Dime] Requirement for the addition of Auditing Functionality toDiameter

[Anders Kristensen <andersk at cisco.com>]

This draft does seem like it could be very useful. Some comments...

The starting point of the draft is that of a rebooted device not knowing 
its state. I wonder if the scope could be expanded to cover failover 
scenarios, e.g. allow a backup node, typically a node belonging to the 
same realm as the failed element, to audit sessions belonging to the 
failed node and take them over (which would amount to the peer 
associating the session with the backup node). This probably raises all 
sorts of issues, some security issues for a start, but could be very useful.

I imagine that for some implementations the validity of a returned 
Query-Index would be time limited -- if the initial query resulted in 
some state being created on the server it will be treated as soft state. 
Does the draft need to talk about that, e.g. have the server indicate in 
the answer how long the Query-Index will be valid or mandate a minimum time?

Should the querying party have a say in how much state is returned in 
one answer?

The peer returns generic (common across apps) and appliation specific 
per-session state. I'm thinking the generic state listed should include 
Class AVPs (cookies) that the peer has stored on behalf of the client 
and also time until the session expires. There are probably other things.

Thanks,
Anders

john.loughney at nokia.com wrote:
> Thanks for the quick reply.  This sounds like a reasonable start - it 
> would be good to get people to read the draft and comment on the general 
> problem.
>  
> thanks,
> John
>
>     ------------------------------------------------------------------------
>     *From:* ext Avri Doria [mailto:avri at acm.org]
>     *Sent:* 22 June, 2006 21:49
>     *To:* Loughney John (Nokia-NRC/Helsinki)
>     *Cc:* hannes.tschofenig at siemens.com; dime at ietf.org; Ulf.Bodin at operax.com
>     *Subject:* Re: [Dime] Requirement for the addition of Auditing
>     Functionality toDiameter
>
>     Hi,
>
>     On 21 jun 2006, at 23.23, <john.loughney at nokia.com
>     <mailto:john.loughney at nokia.com>> <john.loughney at nokia.com
>     <mailto:john.loughney at nokia.com>> wrote:
>
> >     Hi all,
> >      
> >     Hopefully Ulf/Avri are on the list, but I wonder if there is any
> >     overlap to this expired draft:
> >      
> >
> >     __
> >     <http://quimby.gnus.org/internet-drafts/draft-calhoun-diameter-res-mgm>_http://quimby.gnus.org/internet-drafts/draft-calhoun-diameter-res-mgm
> >     <http://quimby.gnus.org/internet-drafts/draft-calhoun-diameter-res-mgmt-08.txt>_
> >     <http://quimby.gnus.org/internet-drafts/draft-calhoun-diameter-res-mgmt-08.txt>
> >     <http://quimby.gnus.org/internet-drafts/draft-calhoun-diameter-res-mgmt-08.txt>t-08.txt
>
>     yes, we definitely aware of the expired draft (and i am on the list
>     not sure about Ulf).
>
>     we are proposing that the requirements in the draft be reviewed and
>     then if the group agrees that the requirements warrant it and when
>     they are complete, that the older draft be reviewed, revived and
>     reworked if appropriate.
>
>     btw, the official id notice:
>     http://www1.ietf.org/mail-archive/web/i-d-announce/current/msg10758.html
>
>
>     a.
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> DiME mailing list
> DiME at ietf.org
> https://www1.ietf.org/mailman/listinfo/dime

_______________________________________________
DiME mailing list
DiME at ietf.org
https://www1.ietf.org/mailman/listinfo/dime