[mnet-devel] "One Hop Lookups for Peer-to-Peer Overlays"

[Some Guy]

 --- Kyle Hasselbacher <kyle-list-mndev at toehold.com> wrote: 

> >> "One Hop Lookups for Peer-to-Peer Overlays"
> >> Anjali Gupta, Barbara Liskov, Rodrigo Rodrigues
> >> 2003 or so
> >> http://www.pmg.lcs.mit.edu/~anjali/papers/onehop.pdf
> >
> >I'm not sure I buy it.  They still want you to notify every node when you
> >leave and enter the system.  Sure you can use a broadcast tree so you
> >don't have to do it, but the bandwidth still has to be used to tell
> >everyone when nodes join and leave.
> 
> If that's true, it's pretty icky.  As the network grows, and join/part
> becomes more common, low bandwidth nodes become saturated just with network
> notifications.  Also, a few flapping nodes can flood everyone.

They estimate that for a system with 10^5 to 10^6 nodes they'll have about 20 to 200 membership
changes a second based on how often people join and leave gnutella and napster.  So how big would
the "here I am" packet have to be?  6B Ip:port, my network ID, maybe a certificate from me, maybe
a certifcate or hash cash or something I use to prove my ID.  It might work.

I think they can solve the flapping node problem with thier broadcast trees.

Still the big problem I have with it is security.


	

	
		
Mit schönen Grüßen von Yahoo! Mail - http://mail.yahoo.de


-------------------------------------------------------
This SF.Net email is sponsored by: Oracle 10g
Get certified on the hottest thing ever to hit the market... Oracle 10g. 
Take an Oracle 10g class now, and we'll give you the exam FREE. 
http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
_______________________________________________
mnet-devel mailing list
mnet-devel at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mnet-devel