Date: Wed, 6 Dec 2000 07:12:09 GMT From: Cliff Sarginson <cliff@raggedclown.net> To: Cliff Sarginson <cliff@raggedclown.net>, Alfred Perlstein <bright@wintelcom.net>, parkpark <parkpark@home.com>, questions@FreeBSD.ORG Subject: Re: theoretical limit for TCP/IP connections Message-ID: <E143YkL-0000Co-00@post.mail.nl.demon.net>
next in thread | raw e-mail | index | archive | help
> > * parkpark <parkpark@home.com> [001205 20:50] wrote: > > > > > > Pleaes Help me to investigate the theoretical limit for TCP/IP > > > connections over a 100Mbps line. I like to know what the line > > > capacity is for a client which does: connect(), write(), close(), > > > and the server does: accept(), read(), close(). These calculations > > > are complicated, but I'm sure soneone can find me some snaswers. > > > Please hemp me to do the calculations as soon as possible as this > > > will give me a better idea of how good my current performance is > > > compared to the theoretical limit for Freebsd!. > > > > Please wrap lines at 70 characters. > > > > Er, I've gotten full 100mbit with not very fast pentium machines. > > > > I'm not sure what you mean exactly by your question though. > > Or do you mean actual connections, doing nothing in particular ? Better clarify ! > I presume what you are really asking is about the efficiency of > the protocol implementations from FBSD. > I guess there are programs out there. But what exactly are you > comparing it with ? > > An easy test would be to do some large file transfers I would have thought. > Maybe some NFS copying. Set up a controlled environment and try it out. > > On the other hand since you have heard of accept,connect and other socket > level system calls you could write your own test program ! That is not > very difficult to do. (if you want to pay me, I will write them for you :). > > > Remember there are a lot of things going on here. > 100Mbits is the hardware spec. of the ethernet, the raw bit transfer rate. > There will also be the TCP/IP overhead. > There will also be the application protocol overhead. > The quality of your network hardware > > Etc. > > Let us know what you are comparing it with... > > Cliff > > > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-questions" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?E143YkL-0000Co-00>