[bittorrent] IO bound

Alan McGovern alan.mcgovern at gmail.com
Fri Apr 13 10:52:09 EDT 2007


> But a HAVE_NOT_ANYMORE message in BitTorrent would solve this and would
> be trivial to implement. (But it would take years untill all clients
> are implementing this .. again ..)


A "have_not_anymore" message wouldn't make sense. I'd assume a client would
only advertise "having" piece if they actually have one. Therefore, once you
have a piece, you have a piece. You want the "HaveNot" so you can control
the pieces another peer can download. This is exactly what the suggest piece
message should be used for.

> This can also lead to reduced throughput, which i believe is one of the
> criticisms of the existing
> > superseeding method using bitfield/have messages.
>
> BT is used to transfer GB of data. Polluting the network with some
> small HAVE messages does way less harm than clients sending broken data
> or unrequested piece (this happens all the time..)


What i actually meant is not that sending all these have's slows down
throughput, it's that you will end up in a situation where Peer X would be
downloading rare pieces from you *IF* you had a full bitfield out, but
because you're doing selective sending of Have Messages, he doesn't know you
have those rare pieces therefore he won't request them off you, thus
reducing throughput.

Alan.

_______________________________________________
> BitTorrent mailing list
> BitTorrent at lists.ibiblio.org
> http://lists.ibiblio.org/mailman/listinfo/bittorrent
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ibiblio.org/pipermail/bittorrent/attachments/20070413/3aa2f2fd/attachment.html 


More information about the BitTorrent mailing list