Date: Mon, 15 Mar 1999 09:13:15 -0600 (CST) From: Jonathan Lemon <jlemon@americantv.com> To: brett@lariat.org, freebsd-chat@freebsd.org Subject: Re: FreeBSD 3.1 SMP outperforms SuSE 6.0 SMP by factor 2.3 !!! Message-ID: <199903151513.JAA17207@free.pcs> In-Reply-To: <local.mail.freebsd-chat/19990315050302$d416@fish.pcs> References: <local.mail.freebsd-chat/19990315001521.A39079@titan.klemm.gtn.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In article <local.mail.freebsd-chat/19990315050302$d416@fish.pcs> you write: >Great! What is the current state of lock pushdown in the FreeBSD >kernel? Have the interrupt handling problems been nuked? The lock pushdown doesn't extend to all subsystems, so certain sections of the kernel are sill serialized, as far as I can tell. On a tangent, this week there is a web proxy bakeoff that is intended to compare performance of high-end web proxies. Of possible interest to this list is that the entire testbed is based on 2.2.7 machines. See http://www.ircache.net, or http//bakeoff.ircache.net/doc/logistics.html Also, at least one of the entries (which I hope will win) is based on a modified 3.1 kernel. Unfortunatly, it's still running a UP kernel since the SMP kernel with this application gives a net loss in performance. Also, in this particular environment, Linux is simply not able to deliver the performance required. Neither is Digital Unix, running on an AlphaServer; it seems they have some, umm, interesting bugs in their TCP/IP stack. -- Jonathan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199903151513.JAA17207>