Re: [savi] A few comments on draft-vogt-savi-framework

[Dong Zhang <zhangdong_rh at huaweisymantec.com>]

Hi Christian,

>Your recollection is correct; the Framework document should and will
>include the properties of, and the relationship between, the various IP
>address assignment methods.  Section 3.1 in the current Framework
>document version is a placeholder for this.  Text for this section is
>still to be added, as noted in the document.  And yes, this text shall
>cover not only SLAAC and DHCP, but also SEND and manual IP address
>configuration.
My concern is that as a vendor, it may implement the savi approaches, 
including fcfs approach, cps approach and SEND approach. So how to coordinate
them needs to be considered. I mean all the approaches work together or choose
one to work. I guess it could be the latter. If so, the condition for choosing 
each approach should be illustrated unambiguously.
However, I assumed that there is still an exception, which means several 
address assignment methods exist in one case. How do the savi approaches deal 
with this case?

Thanks.
------------------				 
Dong Zhang