[p2p-hackers] Using UDT for swarming

Justin Chapweske justin at chapweske.com
Tue Jan 25 21:30:25 UTC 2005


> answers them. One possible exception might be all this FEC vodoo in
> the multicasting group, but for me it feels like a bit of an overkill
> for the job (even if I would be sure that FLUTE/ALC/LCT stack really
> does resolve these issues, which I'm not).

The FLUTE/ALC/LCT stack does indeed solve the long fat network problem
and we have a number of customers deploying the solution for transfer of
extremely large data sets (100 GB+) over both satellite and terrestrial
networks.

However, the protocol stack has a huge learning curve, is quite
complicated to implement, and is massive overkill for many
applications.  

However, our bread-and-butter tends to be providing massive overkill
solutions, so it fits quite nicely with what we do :)  

For the majority of folks, a UDT-type approach might be the best way to
go.  But honestly, UPnP is becoming increasingly deployed, and Joe P2P
Hacker is likely to completely botch basic congestion control, so it
might be best for people to stick with TCP and not potentially bring the
Internet to a screeching halt.

Thanks,

-Justin

P.S. We'll likely be doing a public release of our satellite/multicast
file transfer product sometime this quarter, so anyone that is
interested in the mean time can feel free to contact me directly. 





More information about the P2p-hackers mailing list