[BitTorrent] An alternative "super-seed" mode

Mike Ravkine krypt at mountaincable.net
Sun Jan 30 17:25:35 EST 2005


Jesus Cea wrote:

>Another questions with "super-seed" mode.
>
>1. What is the swarm has only a seeder and a peer?. Nobody is reporting 
>piece propagation, so you never send a new "have" to the peer.
>  
>
This is an edgecase.  I hope the algorithm is smart enough to deal with 
this and fall back to normal seeding.

>2. What if a peer transfer a piece to another peer, but that peer goes
>offline?. The seeder never saw the transfer and penalize the first peer
>not sending it more "have" messages ever. So that peer will be 
>completely ignored.
>  
>
This isn't a problem, as the piece will still eventually show up on 
another peer.

>3. To improve the pipeline effect when there are few peers, I think that
>sending several (let say, 4-10) initial "have" to connecting peers could 
>be appropiate, to maximice download speed. Of course you only give more 
>credit if the propagation of those pieces are confirmed.
>  
>
The goal here is not so much to maximize downloading speed, but to 
minimize the upload overhead before another seed is created.  Speed 
increases from there are what makes BT great.

>4. If the swarm is formed by few peers, you can penalize a high 
>bandwidth peer because the rest of peers are more slow getting the 
>pieces from it.
>  
>
Uhm, again, no.  "High Bandwidth Peers" can upload quickly.  Everyone 
can download quickly.  "High Bandwidth Peers" will enable people to get 
blocks FASTER, since the bottleneck is upload speed 99% of the time.

>As you can see, my concern is with small swarm. Obviously, big swarms 
>have no these problems.
>
>  
>
Generally, small swarms are also in less of a need for superseed.. the 
whole problem superseed is trying to solve is inefficient first seeding 
in a rapidly growing swarm.

>With this environment i'm confused. The original Super-seed algorithm 
>works fine if the swarm is big, but my original proposal seems to do a 
>better work if the swarm is very small (let say, 5 peers).
>
>  
>
What you're trying to do is maximize download speed for every member of 
the swarm. This is NOT THE SAME GOAL as superseed.

>And yes, my working conditions are just that.
>
>Comments, please. Suggestions welcomed.
>
>  
>
What are you efficiency measurements with and without superseed, and 
with your modified superseed?  I'd like to see some numbers..

--kRYPT


 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/BitTorrent/

<*> To unsubscribe from this group, send an email to:
    BitTorrent-unsubscribe at yahoogroups.com

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/
 





More information about the BitTorrent mailing list