[p2p-hackers] Re: scalability

[coderman]

On 11/30/05, Adam Fisk <afisk at speedymail.org> wrote:
> ...
> I was always amazed you were able to make so many connections, although
> I know the eDonkey guys maintain incredibly high numbers of connections
> as well.  That's actually a fascinating finding in light of the paper
> someone mentioned earlier on discovering the optimal number of
> connections to maintain in DHTs -- you might maintain more if you have
> the right connection scheme.  I like the approach of just keeping as
> many connections as you can.

this scalability is only possible because the UDP protocol is designed
for control / messaging communication only.  this means buffer
requirements are minimal and connection overhead very light.  if you
need a bearer transport the congestion avoidance and xmit/recv buffers
become a huge overhead at this scale.


> What ever became of Alpine?

it is hibernating awaiting a rewrite.  i started working on feedbackfs
to improve the nature of implicit feedback on resources obtained via
alpine before starting the rewrite.  feedbackfs got me distracted on
security (given the nature of the details tracked by feedbackfs it
really has to be secure to protect peer privacy) which has been my
current focus.

im recursing back out of this rabbit hole and will resume work on it
at some point.  i'm glad to know that it was useful to others in some
manner; i certainly think the approach has merit regardless of where
it is implemented (GUESS++).