Re: [Dime] [dime] #23: DOIC behavior for realm overload

["Wiehe, Ulrich (NSN - DE/Munich)" <[e-mail]>]

Steve,

see inline.
Ulrich

From: DiME [mailto:[e-mail]] On Behalf Of ext Steve Donovan
Sent: Wednesday, February 19, 2014 9:54 PM
To: [e-mail]
Subject: Re: [Dime] [dime] #23: DOIC behavior for realm overload

We've had a lot of discussion of this topic.

I believe that we reached agreement that we currently have two types of reports:

- Host report that applies to requests sent to a destination-host
- Realm report that applies requests routed to a realm that do not have a specified destination-host (realm-routed requests)

<Ulrich> I agree</Ulrich>

We also have proposed wording on the interaction between these report types.  

<Ulrich> can you please remind me what that proposed wording is. My current understanding is that there is no interaction between these two report types</Ulrich>

I propose that the second be renamed to realm-routed reports.

<Ulrich> no strong view, but my preference is in favour of "absent-host reports"</Ulrich>

A separate ticket has been opened on the need for a third report type that would apply to all request routed to a realm, independent of whether a request contains a destination-host AVP.
<Ulrich> I understand that this is #55</Ulrich>

Steve
On 1/21/14 9:24 AM, dime issue tracker wrote:
#23: DOIC behavior for realm overload

 This applies to draft-ietf-dime-ovli-01, which does not show up in the
 Component drop down menu.

 The current assumption in the -01 draft is inconsistent in the definition
 of behavior of behavior by a reacting node when it receives a realm
 overload report.

 Section 4.6 says overload treatment should apply to all request bound for
 the overloaded realm.

 Section 5.5.2 is not clear and there have been interpretations that a
 realm overload report only applies to requests that contain the realm and
 do not contain a destination-host AVP.

 Section 5.5.2 should be updated to be consistent with section 4.6.