Re: [DNA] Comments on DNA Solution: Link Identifier based approach

[JinHyeock Choi <jinchoe@samsung.com>]

Dear Subba Reddy

Thanks for your kind feedback. 

> > Though that case is not explicitly mentioned, I had that in mind
> > while design the protocol. That's why hosts make no decision
> > when receiving an RA having no linkid prefix. It will work even
> > in such an environment.
> 
> 
> I think it is useful to look into the RA sent by the unmodified router,
> to see if it is advertising the linkid prefix (obviously no I bit is
> available)
> so that host can make a decision that it is still on the same link.

This will work for the most of cases but may cause problem 
w.r.t. flash renumbering and early reassignment, which is 
described in Sub-sec 7.3   Issue 003: 

> Take a case where more than one router boots up, if the routers are not
> replying
> for RSs (not sending RAs) during initial 12 seconds, then they will
> configure different
> linkids and it may take some more time to converge.
> 
> If the link is coming up, routers chosing the linkid based on the prefixes
> it is advertising
> is same as waiting for 12 seconds other than the delay in the later case
> (because routers
> will not come to know each others prefixes)

Eventually the linkid will converge and I expect there will be no 
flapping. I put this as an open issue 006.  

> > The problem with 'the longest valid prefix time' is that there may be
> > multiple prefixes having the same valid prefix time. Then we need
> > an additional decision scheme.
> 
> Even though one more decision is required, if we select linkid
> based on life time, then linkids may not change frequently
> because of life time expiry.

Usually I guess that the most prefixes are configured the default 
valid lifetime, 30 days. Prefixes are not changed so often. So I 
expect it will work fine even if we don't choose the longest lifetime 
prefix. 

Thanks in advance for your kind consideration. 

Best Regards

JinHyeock