Gnutella overview for Web Service Discovery

[M1862]

Hi Guys,

I am new to this the site and am doing a project on Distributed Applications and Web Services and would value your opinions.

Firstly, I have to provide a critical overview of Gnutella I in the context of Web Services discovery and explain TWO advantages and TWO disadvantages in using Gnutella I for Web Services discovery.

I am a little unsure how to apply Gnutella in Web services discovery? What does this actually mean? Isn't p2p Web services discovery?

I would really appreciate any help or advice you could give.

Thanks

M1862

[arne_bab]

Well, Gnutella can do service discovery.

What it can do well are fuzzy searches with many results. Scaling should mostly be O(log(N)), thanks to Dynamic Querying and the Query Routing Protocol (which turns the original pure broadcast model of Gnutelly 0.4 into a very efficient directed search in Gnutella 0.6).

What it can do very well is discovering additional sources of a found service. See the download mesh.

The initial connection in Gnutella actually is service discovery. See UDP Host Caches.

What it cannot do that well are exact searches (more exactly: for exact searches, Gnutella has a DHT - see Mojito. That has guaranteed logarithmic scaling, but can *only* do exact searches efficiently).

Generally see Gnutella for Users - Gnufu

[M1862]

Hi

Thank you for your help, and for pointing me in the right direction. I was a little unsure about which area of Gnutella I should concentrate on, so that's great - Thanks

[arne_bab]

glad to help