Re: [P2PSIP] Re: HIP pros and cons

Recognizing that we are all too stupid from lack of rest to have an 
intelligent conversation this weekend ... but continuing to type anyway ...

:-)

>> - mobility can be achieved by a P2PSIP overlay easily once a P2PSIP 
>> overlay is running. You can just update you current Contact URI. So  here 
>> again, there is no strong argument for HIP
>
> Err.. I would not say that mobility is easy.  However, the ID/Locator 
> split idea of HIP make mobility easier.

I agree with Philip's response here, but want to point to a related topic.

A decent number of the application scenarios would benefit from multihoming, 
and another design goal for Bob Moskowitz-era HIP was supporting IPsec VPNs 
that survived path changes because IPsec only saw HITs, so when the 
underlying IP addresses changed, applications survived.

(That's why HITs look like IPv6 addresses, by the way)

This matters more in my own application scenarios of interest, and the 
ID/Loc split in HIP helps with this as well.

Both mobility and multihoming are discussed (in the same section) in 
http://tools.ietf.org/html/rfc4423.

People may be thinking about SIP-level or overlay-level mobility in their 
environments, but neither of these are attributes in the current version of 
the application scenarios draft.

I think some of the "carrier-grade robustness" attributes could be addressed 
by multihoming, but I'm sure this isn't explicit in the current draft, 
either.

I meant to say this during the talk on routing, but we only had 15 minutes 
and I wanted to give the working group as much precious discussion time as 
possible - BUT,

- I'm not thinking that the application scenarios draft would even be 
adopted as a WG draft, and certainly would not be a prerequisite for 
requirements, concepts descriptions, or peer protocol proposals,

BUT

- It would be nice for people to think about the current draft contents, and 
tell us what they are thinking about, that isn't covered (like mobility, 
like support for survivability).

http://tools.ietf.org/wg/p2psip/draft-bryan-p2psip-app-scenarios-00.txt

If you can provide a column (that describes an application scenario), or a 
row (that does the attribute analysis for multiple application scenarios), 
that would be absolutely lovely.

IMO, of course.

Thanks,

Spencer

_______________________________________________
P2PSIP mailing list
P2PSIP@ietf.org
https://www1.ietf.org/mailman/listinfo/p2psip