Re: [RRG] Mapping Distribution and Management: contribution for comments

[ksriram@nist.gov]

Quoting Brian E Carpenter <brian.e.carpenter@gmail.com>:
> Hi,
> 
> On 2008-07-17 10:09, K. Sriram wrote:
> > I am submitting a contribution (co-authored with Young-Tak Kim and Doug
> > A link to the paper is:
> > http://www.antd.nist.gov/~ksriram/NGRA_map_mgmt.pdf
> 
> I have a couple of questions about Fig. 3:
> 
> Firstly, what is the advantage in the lower layer
> of ETRs (ETR1-6) being ETRs at all? If they are handling
> locators that aggregate at ETR7 and ETR8 respectively,
> why not just en/decapsulate at ETR7 and ETR8?
> 
> (To be clear, the suggestion I put out some months ago didn't
> just have a hierarchy of ETRs; it also had a hierarchy
> of locator domains and a hierarchy of maps, so you would
> decapsulate and re-encapsulate with a different mapping at
> each layer. That was supposed to be a scaling mechanism.)

Thank you for your comments.
We have similar reasons in mind for a hierarchy of ETRs.
Please also see slide #18 in our presentation (draft):
http://www.antd.nist.gov/~ksriram/MDP_Dublin_KS_Slides.pdf
where we mention possibility of recursive map-and-encap.
Further, we are assuming that some subprefixes (subnets) 
are amenable to aggregation at higher layer ETR and yield to 
mapping table reduction while there may be other prefixes 
that are not aggregatable. In the latter case, the prefix
may as well have lower layer ETR as its locator just in 
case there are shorter paths from some ITRs to that ETR.
At this time, we are putting forward some ideas for 
discussion and trying to see where we can go with this. 

> 
> Secondly, your two multihoming examples are only connected
> to second layer ETRs (ETR7 and ETR8). Is there any reason
> you can't multi-home at the first layer?

Sure, they can multi-home at the first layer. What we show 
in the figure was just to state an example. 

Can you please provide me a link to your earlier message
about hierarchy of ETRs?

Thanks again for the comments.
Sriram
> 
>     Brian
> 



--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg