Date: Fri, 26 May 95 13:26:03 MDT From: terry@cs.weber.edu (Terry Lambert) To: phk@ref.tfs.com (Poul-Henning Kamp) Cc: imb@scgt.oz.au, hasty@netcom.com, julian@ref.tfs.com, pss@fore.com, announce@FreeBSD.org, hackers@FreeBSD.org, rv@fore.com Subject: Re: Drivers for FORE systems cards under FreeBSD Message-ID: <9505261926.AA27641@cs.weber.edu> In-Reply-To: <199505261606.JAA01495@ref.tfs.com> from "Poul-Henning Kamp" at May 26, 95 09:06:05 am
next in thread | previous in thread | raw e-mail | index | archive | help
> > Much as I hate generalisations .. I would warn that ATM and IP apparently do > > not mix well in cases where good interactive response is critical, at least > > in my limited experience. I get better response over my own overworked 14k4 > > modem connection to the net than over a (university based) 2 megabit ATM > > link. > > This is not a general problem with ATM, and I tend to say that it's > because there are other users on the 2Mb ATM than you. > > It's certainly not an argument against ATM... Poul's right. It's an argument against bandwidth overcommit. Which is an argument against non-virtual circuit transports for virtual circuit traffic. THAT's the argument against ATM. On the other hand, ATM and Frame Relay source quench doesn't really work well with TCP/IP, since TCP/IP doesn't recognize it as a valid congestion control mechanism. Arguably, the problem is in TCP/IP. Except it came first. Terry Lambert terry@cs.weber.edu --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9505261926.AA27641>