From owner-freebsd-hackers Mon Nov 3 16:28:00 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA16497 for hackers-outgoing; Mon, 3 Nov 1997 16:28:00 -0800 (PST) (envelope-from owner-freebsd-hackers) Received: from gatekeeper.tsc.tdk.com (root@gatekeeper.tsc.tdk.com [207.113.159.21]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id QAA16492 for ; Mon, 3 Nov 1997 16:27:58 -0800 (PST) (envelope-from gdonl@tsc.tdk.com) Received: from sunrise.gv.tsc.tdk.com (root@sunrise.gv.tsc.tdk.com [192.168.241.191]) by gatekeeper.tsc.tdk.com (8.8.4/8.8.4) with ESMTP id QAA17616; Mon, 3 Nov 1997 16:27:32 -0800 (PST) Received: from salsa.gv.tsc.tdk.com (salsa.gv.tsc.tdk.com [192.168.241.194]) by sunrise.gv.tsc.tdk.com (8.8.5/8.8.5) with ESMTP id QAA23859; Mon, 3 Nov 1997 16:27:31 -0800 (PST) Received: (from gdonl@localhost) by salsa.gv.tsc.tdk.com (8.8.5/8.8.5) id QAA10243; Mon, 3 Nov 1997 16:27:29 -0800 (PST) From: Don Lewis Message-Id: <199711040027.QAA10243@salsa.gv.tsc.tdk.com> Date: Mon, 3 Nov 1997 16:27:29 -0800 In-Reply-To: Charles Henrich "Re: de0 errors" (Nov 3, 3:29pm) X-Mailer: Mail User's Shell (7.2.6 alpha(3) 7/19/95) To: Charles Henrich , Don Lewis Subject: Re: de0 errors Cc: Matt Thomas , Jaye Mathisen , freebsd-hackers@FreeBSD.ORG Sender: owner-freebsd-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Nov 3, 3:29pm, Charles Henrich wrote: } Subject: Re: de0 errors } On the subject of Re: de0 errors, Don Lewis stated: } } > A PCI bus analyzer could be useful if the problem is what I think it is. } > What processor and PCI chipset do you have? } } This is an ALR Revolution MP-Pro system w/ 1 200Mhz PPro, 256MB Ram, 4 SCSI } Cards, a Video card, and a SMC Etherpower 10/100 Card. Its an Intel 440FX } chipset.. Bingo! I believe that a 21140 based NIC is bad juju in a system with a 440FX based system because the 21140 can only do plain PCI memory reads (MR) and you really need to use PCI read multiple commands (MRM) to get decent throughput with this chipset, and probably any other Pentium Pro or Pentium II chipset. A 21140A based NIC should work much better since it can use the MRM command. } > > Just using FTP between two hosts on the 100mbit net I see 2-3MB/sec and } > > receive at about half that. } > } > That's not very impressive. To eliminate the influence of disk I/O, you } > might want to test network throughput using ttcp or something similar. I'd } > recommend using ping to check for packet loss at the same time. You } > shouldn't see any packet loss if you're sending between two machines on the } > same network segment. } > } } Humm, tcpblast shows 4.0MB/sec. Granted this isnt an empty net either, fairly } busy I would imagine. That's probably a reasonable number for a 10Mb network, but not a 100Mb network. BTW, I bet this NIC is also hurting your disk I/O throughput since it is hogging the PCI bus because it's using an inefficient transfer method. --- Truck