Re: Using the DHT to find download sources

--- In [email protected], Raphael_Manfredi@... wrote:
>
> Quoting pwang708 <wang.peng@...> from ml.gnutella.dev-forum:
> :Michael and I discussed grouping peers having common interests. We can
> :also somehow tag the groups, then publish the tags and a subset of
> :current ultrapeers in the group into the DHT. 
> 
> Such clustering is a very bad idea.
> 
> In the end, it makes it very difficult for people outside the cluster
> to find something from within the cluster, and it can also create an
> island in the network if the clustering is too strong.
> 
> Also, clustering along dimensions such as interets or language in the
> hope that QRP routing will be much more useful for your searches, is a
> double-edged sword: it can make it impossible for you to find something
> your cluster is not sharing, because the nodes that do share the things
> you now want are just a few hops away past your clustered horizon.
> 

Clustering will change the topology of the network. And we should be
careful. Large scale simulation is needed before deploy if we are
going to do this.

It's true that having a strong clustering may result in islands in the
Gnutella network. And one may have trouble to find things its Gnutella
clusters do not have. We do have the DHT, however, which will help us
to "jump" to the clusters having the files that we cannot find in our
clusters. In addition, we can always reserve some connections, say
5~10 out of 30, to random ultrapeers in the Gnutella network to
improve connectivity between clusters. 


Best,
Peng