Kevin Atkinson:
> 1) Is there any way to determine the amount of extra load queries for non 
> existent data put on the network and the cause of those queries (ie DBR's, 
> version sites loading images that don't exist, active link site loading 
> images that don't exist, etc.)

Run a node, observe the failure/success ratio.

> 2) Is there a way to get statistics for the number of nodes that data goes 
> through for successful queries?

You can estimate based on latency.

> 3) Exactly how does freenet determine which node to contact to pass on a 
> query?  Fell free to point to easy to read, well documented code, if that 
> will be easier.  I have not actually looked at the freenet source so I 
> have no idea what its like.  If it is not easy to tell from the code 
> please describe it here.

Numerical key distance.

> 4) Are there any programs to simulate how a bug free freenet, or 
> similar networks will perform under various types of loads and network 
> configurations.  If so could some one point me to such a simulation and in 
> particular the results for running it on a freenet type network.

Serapis and Aurora, in CVS. Look at the papers and the P2P book for
results.

> 5) Now for the big one.  What is currently wrong with freenet now.  I am 
> more interested in performance issues than I am with limitations to what it 
> can do.  But I am willing to listen to both.  You are more than welcome to 
> point me to past mailings (or better yet post summaries) and the like.

In implementation or design?

_______________________________________________
freenet-tech mailing list
[EMAIL PROTECTED]
http://lists.freenetproject.org/mailman/listinfo/tech

Reply via email to