On Sun, Jan 26, 2003 at 01:00:01PM -0500, Gianni Johansson wrote: > On Saturday 25 January 2003 22:49, you wrote: > > > length 7.4M > > > elapsed time: 37 minutes > > > end to end transfer rate: 7.4M * (1024K/M) / (37 min * (60sec/min) ) ~= > > > 3.4k/sec > > > > > > It took one minute to find the SplitFile metadata and somewhat less than > > > 2 minutes to decode. > > > > > > Discounting that ~= 3.7k/sec raw transfer. > > > > > > How many threads do you use to get 30k/sec? > > > > The default. Perhaps the initial retrieval time is more significant > > with a smaller file (the one I downloaded was ~200MB). I will do some > > more tests. > > > I think the larger issue is that you are comparing a download of a file that > is well known and often requested to one that is relatively unknown. > > If you want to do streaming you have to be able to get the *unknown* file at > a reasonable transfer rate. This is an excellent point. However, the stream would surely be inserted at a reasonable HTL? Propagation shouldn't affect bandwidth once we have started a transfer, so it all comes down to how much parallelism we can use... which depends on the length of the buffer, and the number of FEC chunks each block is divided into... > > I would be very interested in seeing the results of your experiments. Definitely. > > --gj > > > Ian. >
-- Matthew Toseland toad at amphibian.dyndns.org/amphibian at users.sourceforge.net Full time freenet hacker. http://freenetproject.org/ Freenet Distribution Node (temporary) at http://amphibian.dyndns.org:8889/RMMHBxLOISw/ ICTHUS. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20030127/9149dd95/attachment.pgp>