Re: identity persistence and comparison issues

[marcelo bagnulo braun <mbagnulo@ing.uc3m.es>]

Hi Joe,

El 25/06/2004, a las 2:39, Joe Touch escribió:
> We do know. The lifetime is unlimited. A TCP connection can stay up as 
> long as the two endpoints stay up.
>
> The identifier needs to be stable at the endpoints as long as the 
> endpoints care - which is why I like how HIP negotiates these labels, 
> and maintains the mapping at the endpoints.

Yes, but for how long does the hip endpoint maintains the mapping 
between the HIT and the locators of the other endpoints?

I mean, there must be a garbage collection mechanism, if not the hip 
endpoint will have to store infromation about all the endpoints that it 
has communicated with in the past

regards, marcelo

>  The endpoints are the only place that can or should know about 
> whether they change.
>
> Joe