I never talked about blocking (no one will accept it)
but either controling (ie limiting or giving priority
to everything but P2P) or fairness per user (without
considering the application). From my experience 5-10%
of users are consuming 70-80% of network ressources...
all paying the same price so not fair!!!.


--- "Pablo A. Destefanis (Ecoband)"
<pablo@...> wrote:
> Ok Patrick, you're right. But with the popularity it
> has achieved, I
> don't think blocking is a solution. Maybe a
> combination of local caching
> with some policing could be. If the statistics I've
> seen so far are
> right, then the caching works wonders. Imagine what
> that could be for
> your users! 
> 
> I guess that if you block (specially for residential
> users) they will
> flee to some other provider. Also blocking affects
> P2P as a system...
> (which I like, but that's a personal opinion ;-)
> 
> Pablo.
> 
> -----Original Message-----
> From: patrick deroudilhe [mailto:pderoudi@...> 
> Sent: Thursday, July 10, 2003 10:52
> To: isp-caching@isp-caching.com
> Subject: [isp-caching] RE: interresting
> 
> 
> I do agree but maybe many of your customers are
> suffering because a limited number of them are
> aggressively consuming YOUR bandwidth (lets consider
> it is not unlimited)by starting large number of
> concurrent connections. 
> That why Allot is also proposing a different
> approach
> to control P2P: fairness per users (and not per
> connections. No limitation - no application control
> -
> just equity per user when your bandwidth is
> congested.
> 
> If you are concerned by doing "any effort to improve
> users experience" it could be a good complement !!
>  
> 
> --- Amos Rosenboim <r.amos@....il> wrote:
> > My own two cents:
> > 
> > P2p is one of the great things that happened in
> the
> > internet.
> > It really follows the model of a great open public
> > network.
> > I don't want to limit or shape my users traffic
> > because I believe they
> > are entiteled to do what ever they want with the
> > bandwidth they are
> > paying for.
> > But since p2p is challanging my network I will do
> > any effort to improve
> > my users experience using caching/redirection like
> > the solutions from
> > expand/joltid/sandvine
> > 
> > Regards
> > Amos
> > 
> > 
> > -----Original Message-----
> > From: patrick deroudilhe
> [mailto:pderoudi@...> > 
> > Sent: Thursday, July 10, 2003 12:32 PM
> > To: isp-caching@isp-caching.com
> > Subject: [isp-caching] RE: interresting
> > 
> > Question was :
> > 
> > > Interesting indeed! Is anyone using any caching
> > engine to catch P2P traffic? Are there other
> > products
> > out there?
> > 
> > You can also use Traffic shappers to control P2P
> > bandwidth allocation (I aggree it is different
> > approach).
> > www.allot.com is doing nice P2P
> > (KazaaV2-Gnutella)application recognition
> > (signature)
> > and shapping up to 1Gbps throughput. They also
> > propose
> > Transparent Cache Redirection option.
> > 
> > Regards
> > Patrick
> > --- "Pablo A. Destefanis (Ecoband)"
> > <pablo@...> wrote:
> > > Thank you Amos and everyone.
> > > 
> > > BTW, there's a pretty good white paper on
> caching
> > > P2P traffic that was
> > > written by Nathaniel Leibowitz, Aviv Bergman and
> > > others based on some
> > > very detailed tests they did with Amos (I think
> > so,
> > > am I right?).
> > > 
> > > Very informative.
> > > (http://2002.iwcw.org/papers/18500253.pdf)
> > > 
> > > Does anyone ever tried to use iptables to do the
> > L4
> > > switching on a
> > > multihomed Linux box? I don't have the budget
> for
> > an
> > > Alteon box... ;-)
> > > 
> > > Cheers,
> > > 
> > > Pablo.
> > > 
> > > -----Original Message-----
> > > From: Amos Rosenboim
> [mailto:r.amos@...> > > Sent: Thursday, July 10, 2003 08:38
> > > To: isp-caching@isp-caching.com
> > > Subject: [isp-caching] RE: interresting
> > > 
> > > 
> > > The first p2p cache that i know of was developed
> > by
> > > expand networks from
> > > israel. I have two of 3 of those in my network
> for
> > > fastrack ( Kazza )
> > > traffic. It used to be very effective when kazza
> > was
> > > working on a single
> > > tcp port (1214). now we need to do all kinds of
> > > layer 4 -7 tricks to get
> > > the traffic to the cache. Every cache of mine
> > deals
> > > with roughly
> > > 100Mbits/sec. Expand networks stopped the
> > > development of their product,
> > > however, since then there are new products like
> > > mushrooms. the examples
> > > i know of:
> > > 
> > > Joltid - the same people who are behind kazza,
> > they
> > > have a very nice
> > > product.
> > > 
> > > 
> > > Sandvine - more complicated product then most of
> > > them but a vey serious
> > > compnay.
> > > 
> > > Regards
> > > Amos
> > > 
> > > 
> > > -----Original Message-----
> > > From: Moshe Konortov [mailto:moshek@...> > > Sent: Thursday, July 10, 2003 10:28 AM
> > > To: isp-caching@isp-caching.com
> > > Subject: [isp-caching] RE: interresting
> > > 
> > > I'm doing it as a reseller for those product and
> > > also Expand doing it
> > > 
> > > Regards,
> > > 
> > > Moshe Konortov
> > > 
> > > -----Original Message-----
> > > From: Mark Pace Balzan
> > [mailto:mpb@...> > > Sent: Thursday, July 10, 2003 8:28 AM
> > > To: isp-caching@isp-caching.com
> > > Subject: [isp-caching] RE: interresting
> > > 
> > > 
> > > Has anyone been using/testing these products
> > > 
> > > care to comment about their performance and
> > > effectiveness.
> > > 
> > > 
> > > cheers
> > > 
> > > Mark
> > > 
> > > 
> > > ----- Original Message -----
> > > From: "Moshe Konortov" <moshek@ankor.co.il>
> > > To: <isp-caching@isp-caching.com>
> > > Sent: Thursday, July 10, 2003 8:39 AM
> > > Subject: [isp-caching] RE: interresting
> > > 
> 
=== message truncated ===


__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com

------------------------ANNOUNCEMENT---------------------------------
     *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-
                         ** 802.11 Planet **
       The business of hot spots. Wireless ISP business plans.
      Wireless network certifications. High-gain antenna design.
                       Site survey techniques.
                       June 25 - 27, Boston, MA
             http://www.jupiterevents.com//80211/spring03
     *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-

Replies

[isp-caching] RE: interresting, Pablo A. Destefanis \(Ecoband\)