Date: Fri, 17 May 1996 21:39:08 -0600 From: Nate Williams <nate@sri.MT.net> To: Michael Smith <msmith@atrad.adelaide.edu.au> Cc: nate@sri.MT.net (Nate Williams), hackers@FreeBSD.org Subject: Re: Note from Usenet Message-ID: <199605180339.VAA00514@rocky.sri.MT.net> In-Reply-To: <199605180220.LAA15853@genesis.atrad.adelaide.edu.au> References: <199605171652.KAA27616@rocky.sri.MT.net> <199605180220.LAA15853@genesis.atrad.adelaide.edu.au>
next in thread | previous in thread | raw e-mail | index | archive | help
> > Gleaned froma Linux mailing list. > > > > > About a week ago, Alan Cox commented that the networking code in 1.3.xx > > > wasn't up to coping with the type of load that a WWW server would > > > experience, and that 1.2.xx was a better choice for that application. > > > > Apparently Linux 1.3 with the re-written TCP/IP code still isn't up to > > the task. > > Sheesh. What it must have cost him to say _that_. Bit of a pity, really. And he's the author of the new code, so what he says is most definitely an unbiased opinion. I saw some other notes on other articles that it appears to be a problem with the new kernel malloc routines. Apparently the real-time extensions are getting in the way of the kernel being able to service it's own internal memory requests for things like network buffers and such. Nate
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199605180339.VAA00514>