Re: Re: Using the DHT to find download sources

Am Freitag 25 Juli 2008 21:13:27 schrieb pwang708:
> So the query routing table protocol will be inefficient because we
> cannot cluster hashes, i.e., pushing hashes out for two hops is
> expensive. To make things clear, I did some calculation. If we apply
> query routing table protocol to hashes, then an UP needs to spend in
> the order of 1~10 KB per second. (A hash is 20 bytes long. Assuming
> every client shares 1000 files, a leaf connect to 5 UPs, each UP has
> 30 leaves, 0.5M UPs in the network, and 20M session are launched per
> day.) We can do some optimization and bring the cost down below 1 KB
> per second, or even less. Now the question is "does it worth it?"

[...]

> But hash 
> searches may still be expensive and Query routing table protocol has a
> non-trivial cost itself.

Jupp, and using Hashes that cost would be higher. 

> In summary, hash searches in Gnutella is most likely expensive unless
> we can make hits happen within couple of hops. This "unless..." part
> is an interesting topic to think about. :-) Peer group may help...

That's why a DHT might be vastly more efficient for hash searches. 

It doesn't have to return all results for a hash. A few recent sources 
suffice, the Gnutella download mesh will help retrieving the rest of the 
sources. 

Then a high-level representation of Gnutella could be: 

Gnutella: 
- Bootstrapping - Internal Host-Cache and UDP host caches
- Keyword query network: Gnutella 0.6 searches
- Hash query network: Kademlia/Mojito
- Distributed data replication: Download Mesh, FW2FW, ...

Best wishes, 
Arne


[Non-text portions of this message have been removed]